Emoticons and emojis in cross-cultural perspective: Narrative systematic review - IS MUNI
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
FAKULTA SOCIÁLNÍCH STUDIÍ Emoticons and emojis in cross-cultural perspective: Narrative systematic review Bakalářská práce JIŘÍ SVOBODA Vedoucí práce: Mgr. Lenka Dědková, Ph.D. Katedra mediálních studií a žurnalistiky Program Mediální studia a žurnalistika Brno 2022
EMOTICONS AND EMOJIS IN CROSS-CULTURAL PERSPECTIVE: NARRATIVE SYSTEMATIC REVIEW Bibliografický záznam Autor: Jiří Svoboda Fakulta sociálních studií Masarykova univerzita Katedra mediálních studií a žurnalistiky Název práce: Emoticons and emojis in cross-cultural perspective: Narrative systematic review Studijní program: Mediální studia a žurnalistika Vedoucí práce: Mgr. Lenka Dědková, Ph.D. Rok: 2022 Počet stran 68 Klíčová slova: emoji, emotikon, počítačově mediovaná komunikace, systematická přehledová stať, mezikulturní rozdíly, nonverbální komunikace 2
EMOTICONS AND EMOJIS IN CROSS-CULTURAL PERSPECTIVE: NARRATIVE SYSTEMATIC REVIEW Bibliographic record Author: Jiří Svoboda Faculty of Social Studies Masaryk University Department of Media Studies and Journalism Title of Thesis: Emoji usage in cultural perspective – a systematic review Degree Programme: Media Studies and Journalism Supervisor: Mgr. Lenka Dědková, Ph.D. Year: 2022 Number of Pages 68 Keywords: emoji, emoticon, computer-mediated communication, systematic review, cross-culture differences, nonverbal communication 3
EMOTICONS AND EMOJIS IN CROSS-CULTURAL PERSPECTIVE: NARRATIVE SYSTEMATIC REVIEW Anotace Emoji a emotikony jsou zásadní součástí každodenní digitální komuni- kace a vstupují i do interakcí mezi lidmi z různých kultur. Tato bakalář- ská práce si proto klade za cíl prozkoumat tyto nonverbální komuni- kační atributy v jejich mezikulturní perspektivě. Využívá formy syste- matické přehledové statě dvaceti studií z let 2006–2021, které dohro- mady akumulují poznatky z milionů případů použití emoji a emotikonů. Jejich syntéza ukazuje, že dosavadní vědecké, multidisciplinární po- znatky naznačují rozdíly ve vzorcích užívání emoji a emotikon mezi de- sítkami kulturních celků. Zatímco emotikony jsou užívané a chápané ve zřejmých odlišnostech mezi východem a západem, emoji vykazují znaky větší míry univerzálního užívaní. 4
EMOTICONS AND EMOJIS IN CROSS-CULTURAL PERSPECTIVE: NARRATIVE SYSTEMATIC REVIEW Abstract Emojis and emoticons are essential in everyday digital communication and enter into interactions between people from different cultures. This bachelor thesis aims to explore these nonverbal communication attrib- utes in their cross-cultural perspective. It uses the form of a systematic review of twenty studies from 2006-2021, which together accumulate findings from millions of cases of emoji and emoticon use. Their synthe- sis shows that existing scholarly, multidisciplinary findings suggest dif- ferences in patterns of emoji and emoticon use across dozens of cultural units. While emoticons are used and understood in distinctly different ways between East and West, emoji show signs of a greater degree of universal use. 5
EMOTICONS AND EMOJIS IN CROSS-CULTURAL PERSPECTIVE: NARRATIVE SYSTEMATIC REVIEW Declaration Prohlašuji, že jsem bakalářskou práci na téma Emoticons and emojis in cross-cultural perspective: Narrative systematic review zpraco- val sám. Veškeré prameny a zdroje informací, které jsem použil k sepsání této práce, byly citovány v textu a jsou uvedeny v seznamu použitých pramenů a literatury. Brno January 1, 2022 ....................................... Jiří Svoboda 7
EMOTICONS AND EMOJIS IN CROSS-CULTURAL PERSPECTIVE: NARRATIVE SYSTEMATIC REVIEW Acknowledgments I would like to thank Dr. Lenka Dědková for her patience, months of collaboration, and hundreds of insightful comments on earlier versions of this thesis. I would also like to thank Dr. Michal Tkaczyk for the op- portunity to prepare for this thesis in the Introduction to Systematic Literature Review course. Last but not least, immense gratitude goes to my family and friends, who for more than a year have tolerated my con- stant need to share fun facts about emojis and emoticons and have in- spired me to write various parts of this thesis. Šablona DP 3.2.2-FSS-anglicky (2021-04-29) © 2014, 2016, 2018–2021 Masarykova univerzita 9
TABLE OF CONTENTS Table of Contents List of Figures 13 List of Tables 14 1 Introduction 15 2 Emoticons and emojis 17 2.1 DEFINITIONS 17 2.2 HISTORY OF EMOTICONS AND EMOJI 18 3 Theories related to emoticons and emojis 23 3.1 MEDIA AND COMMUNICATION THEORIES 23 3.2 SOCIAL AND PSYCHOLOGICAL THEORIES 24 3.3 CROSS-CULTURAL THEORIES 26 4 Emoticon and emoji as a research subject 30 4.1 CULTURAL PREDICTORS OF EMOJI AND EMOTICONS 31 4.2 INDIVIDUAL PREDICTORS OF EMOJI AND EMOTICONS 33 4.3 EMOJI AS AN OBJECTIVE RESEARCH TOOL 35 4.4 RESEARCH QUESTION 35 5 Method 37 6 Results 42 6.1 EMOTICONS IN CULTURAL PERSPECTIVE 42 6.2 EMOJIS IN CULTURAL PERSPECTIVE 44 6.3 USED THEORETICAL FRAMEWORKS 46 6.4 OTHER OBSERVATIONS 47 7 Discussion 48 8 Conclusion 52 Bibliography 54 Appendix A Research table 65 11
LIST OF FIGURES List of Figures Figure 1: Le point d'amour, as proposed by Hervé Bazin (p. 19) Figure 2: Flow diagram of the systematic review (p. 39) 13
LIST OF TABLES List of Tables Table 1: Break down of numbers of records found by database (p. 38 Table 2: Key emoticon findings (p. 43) Table 3: Key emoji findings (p. 45) 14
1 Introduction Emoticons and emojis possess substantial parts of today’s interpersonal computer-mediated communication (CMC). Billions (Evans 2017) are sent, viewed, and somehow interpreted daily by people worldwide. Since the 1980s, when they first became present in digital textual com- munication, emoticons and emojis largely impacted the internet and communication services – up to the point when half of all comments on Instagram included an emoji (Instagram, 2015). Such widespread usage of these enriching aspects of textual communi- cation naturally generated interest among scholars. Up to October 2021, at least hundreds of studies tried to examine emojis’ and emoti- cons’ design, functions, effects, or interpretations by different groups of users. For example, while seems to be an innocent emoji at first glance, Thomson (2018) found out that up to 34 % of responders use it in a sexual context. Another example would be the emoji used in American English to signal “goodbye.” In British English, it is often as- sociated with traveling and utilized together with emojis of flags (Barb- ieri et al., 2016a). Part of the research on emoticons and emojis is focused on differences in their usage between bigger social groups, such as nations, speakers of a particular language, or culture. While the design of emoticons tend- ed to be culturally dependent from the very beginning (see chapter 2.2.), emojis, on the other hand, are unified. However, that does not mean that all the people worldwide would use them in the same way. For instance, one study found substantial differences in usage of emojis between western and eastern cultures (Guntuku et al., 2019). Still, vari- ations have been found between individual countries (SwiftKey, 2015) or even cities within one country (Barbieri et al., 2016b). Some studies also linked emojis and emoticon usage with cross-cultural indexes (Hall 1960, Hofstede 2011). For example, users in individualistic cultures are expected to express their emotions more than users in collectivistic cultures (Nisbett & Masuda, 2003), hence also having the potential to use more emojis and emoticons in general, as they are used generally to express some kind of emotion. 15
As society is becoming more globalized, digital communication is hap- pening more frequently between members of different cultures. While the language barrier is the main problem in such cases, nonverbal signs also play a substantial role like in any other form of communication, and they too can be culture-dependent. In those cases, diverse under- standings of emojis and emoticons can result in misunderstandings but, on the other hand, can unite the users if the meaning of a given emoji is the same between cultures (see Meyer, 2014). Examining the usage of emoticons and emojis thus represents a great opportunity to learn more about the cultural differences (or lack thereof) in interpersonal communication patterns. Since these aspects are widely spread digital- ly, there is also a lot of objective data available, such as from social me- dia. Therefore, this thesis explores cultural differences in CMC, related explicitly to emoticon and emoji usage. Digital communication is, by default, a multidisciplinary academic field (Šmahel & Černíková, 2019). As research of emojis and emoticons is part of it, underlying theories and critical findings come from media studies, psychology, sociology, linguistics, informatics, human- computer interaction, and in the case of my focus, also from cross- cultural studies. A large scope of fundaments possibly contributed to the fact that little effort has been made to synthesize findings in emoji and emoticon research. To the best of my knowledge, no action has been taken to synthesize their findings in cultural contexts. So how ex- actly do emoticons and emojis usage differ between cultures? Can varia- tions in terms of frequency, interpretation, or context be observed? And if so, can these variations be attributed to some known cultural specif- ics? This thesis tries to answer these questions via a systematic review of studies that examined differences in usage patterns of emojis and emoticons in cultures, nations, or regions of the world. 16
2 Emoticons and emojis 2.1 Definitions This thesis deals mainly with two aspects of computer-mediated com- munication (CMC) – emoticons and emojis. Firstly, an emoticon (plural: emoticons) is “an image made up of symbols such as punctuation marks, used in text messages, emails, etc. to express a particular emotion” (Cam- bridge Dictionary, 2021b). In other words, they are produced, invented, and altered mainly by the users themselves by synthesis of different characters. Examples include :-), :( , -_- or ¯\_(ツ)_/¯, and as the follow- ing chapter will show, users developed distinctive styles of emoticons over the years. Emoticons are present in today’s CMC but not prevalent. According to McCulloch (2020, p. 268), the emoji “face with tears of joy” replaced the most popular emoticon :) in usage levels already in 2015. Secondly, an emoji (plural: emoji or emojis) is “a digital image that is added to a message in electronic communication in order to express a particular idea or feeling” (Cambridge Dictionary 2021a). Among exam- ples are , , , or . As opposed to emoticons, they became pop- ular later (see chapter 2.2.) and, on a technical level, are encoded in CMC as individual characters (same as ڨ, @, ý or 韓), not a set of them. The emoji system1 is also standardized by Unicode Consortium, which maintains the Unicode – the norm that ensures every character is dis- played and understood similarly across devices, no matter the operat- ing system or font. The significance of emojis in today’s digital communication can hardly be overstated. However, the exact usage scale is hard to measure since emojis are now integral parts of almost all digital communication tools, including work-related ones (see Slack.com, Café (app.at.cafe), etc.). Already back in 2015, Instagram revealed that over half of all the com- ments on the platform contained an emoji (Instagram, 2015). Evans 1 From a language perspective, a word emoji can either describe the individual symbol or the system of emoji as a whole 17
(2017) wrote that people send over 6 billion emojis daily. Still, the same year Messenger, a service by Meta, revealed that users only on this plat- form use over 5 billion emojis a day (Facebook Messenger, 2017), so today, the total number may be even higher. To the best of my knowledge, there are no such statistics for the usage of emoticons. This can be attributed to the endless number of emoticon varieties (as op- posed to unified emojis), hence their tracking is more complicated. 2.2 History of emoticons and emoji A crucial reason for using emoticons and emojis is that text-based communication can be seen as limiting as words may not convey the desired meaning properly (Holtgraves & Robinson, 2020). Although I will focus solely on computer-mediated communication in my system- atic review, such limitations of text-based communication apply to oth- er media (e.g., print media, letters). People, therefore, came up with var- ious proposals to solve this issue. For example, Henry Denham, a 16th century English printer, came up with – ؟a symbol he called “a percontation point” to signify rhetorical questions. A 17th-century English philosopher, John Wilkins, proposed a specific punctuation mark to indicate irony. The symbol he chose was ¡, today’s upside-down exclamation point. Later, a French novelist Her- vé Bazin invented five different punctuation marks for various situa- tions, including one visually close to today’s emoticons. “Le point d’amour” (see Figure 1) consisted of two opposing question marks with a common point under them and was supposed to signify love (Sear- geant, 2019). Therefore, Bazin’s invention is essentially a historical ver- sion of today’s ❤or
Figure 1: Le point d'amour, as proposed by Hervé Bazin The need to convey new dimensions of meaning remained present even in the digital era, with emoticons. There is a consensus (see Seargeant, 2019 or McCulloch, 2020) that their beginnings can be traced to Carne- gie Mellon University (CMU) computer messaging system in 1982. Mainly its content consisted of serious announcements from campus, but some students started posting jokes. A professor named Scott Fahlman then wrote the following sequence (Carnegie Mellon Universi- ty, 1982): 19-Sep-82 11:44 Scott E Fahlman :-) From: Scott E Fahlman I propose that the following character sequence for joke mar- kers: :-) Read it sideways. It is probably more economical to mark things that are NOT jokes, given current trends. For this, use :-( This message is most likely the first instance in which an emoticon was used. Other participants soon followed, creating their emoticons, too as some of the messages show (Carnegie Mellon University, 1982): (:-) for messages dealing with bicycle helmets @= for messages dealing with nuclear war
oo for somebody’s headlights are on messages o>-
pan (for example, Apple wanted Japanese users to buy iPhones), in 2010, the first set of emojis was incorporated as part of Unicode 6.0. (McCulloch, 2020, p. 267; Davis & Holbrook, 2021) A year later, in 2011, Apple was the first company to include emojis in their keyboard on the iOS operational system. (Apple Inc., 2011) and in 2013, Google followed with support on Android (Android, 2013). Emojis gained traction over the years, and in 2015 several significant events occurred that seemed to push them into the mainstream. First, emojis were supported in every internet browser (Evans, 2017, p.11). Secondly, as selected by Oxford Dictionaries, the word of the year was “face with tears of joy.” Jurors argued that this very emoji was global- ly the most used in 2015 and had seen a sharp rise from the previous year. They also observed the same trend with the word emoji itself (Ox- ford Languages, 2015). Several social phenomena can explain the rise of emojis. McCulloch (2020, p. 270) argued that they strike a medium between emoticons that can be seen as too simplistic and GIFs that are too complex. Evans (2017, p. 22) attributes it to the global adoption of both social media and smartphones, which saw a sharp rise in the 2010s, as smartphone keyboards were among the first tools globalized for emojis. In 2018, it was estimated that more than 5 billion people worldwide had a mobile phone, with over half of them being smartphones (Silver, 2019). Simi- larly, for example, services by Meta (formerly Facebook) which all sup- port emojis, are used today by more than 3 billion people who send over 100 billion messages every day (Meta, 2021). Emojis also entered actual life events, such as when in 2016, a 12-year- old in Virginia, US, was faced with threatening charges for using sequence online2 (The Washington Post, 2016). Since then, there have been more essential landmarks in emoji adoption, as summarized by Emojitimeline.com (Hånberg Alonso, 2021). Every year, the Unicode Consortium releases a new set of emojis, the most recent being Emoji 14.0, which brings the total number of emojis in the Unicode Standard 2 At the time, the pistol emoji looked like an actual revolver, not a water pistol. Apple introduced the change later in 2016 and other platforms soon followed. 21
to 3,633 (Emojipedia FAQ, 2021). Some still cause controversies, such as the recently revealed emoji of a pregnant man (Murphy Kelly, 2021). Interestingly, members of some nations think that today’s emojis are not sufficient for some culturally dependent communication. Therefore, Sadiq and Shahida (2019) reacted by presenting Pakistani emoji. Fin- land also claims to be the first country to design its national emoji, in- cluding a flag of Åland islands, Nokia phone, or aurora borealis (This is Finland, 2019). None of them, however, are available in Unicode. Therefore, the development of emojis and emoticons in CMC (from the 1980s) shows consistent demand for them. Over time, emojis became common beyond purely messaging platforms and are almost omnipres- ent in our daily lives. As the following chapters will show, such demand also sparked interest among researchers from media, communication, psychology, or cross-cultural studies. 22
3 Theories related to emoticons and emojis Before asking specific questions about cultural differences in emoticons and emojis, we have to see why they are used so widely in the first place. The emoticon and emoji research that has been conducted so far is mostly interdisciplinary, hence it is based on many theories and models from fields of media, linguistics, communication, psychology, and cultural studies. This chapter seeks to briefly introduce the most relevant ones for my thesis. 3.1 Media and communication theories Several mass media and communication theories are used in CMC and emoticon and emojis research. For instance, Daft and Lengel (1986) came up with the media richness theory. In its basics, it says the richest form of communication is face-to-face because it has the largest number of non-verbal cues; it is not mediated but instant and allows personali- zation. On the other hand, purely textual CMC is, in this framework, less rich because people depend purely on words, without any nonverbal cues, and communicate in an asynchronous and mediated environment. Emojis and emoticons can be therefore seen as aspects that enrich CMC. While highly influential in its time, Daft’s and Lengel’s theory has also been criticized. Because it was developed at the early age of CMC and mainly applied to mass media, it can hardly capture the specifics of to- day’s, for example, social media that are interactive and provide more information about each communicator than pure textual message one sends. Walther (2011) argued that media richness theory does not ap- ply to media that allow a wide variety of usage. In other words, Daft’s and Lengel’s approach does not consider the personal factors of partici- pants in CMC and focuses on objective facts of a medium. Further theories, therefore, considered such individual factors. Walther (1992) came up with the social information processing theory. It sees a person receiving a message as an active element in the communication process who can decode the meaning sender intended. This can be done with cues like grammar, punctuation, sentence conjugation, emoticons, and emojis. However, the theory presumes that the communicators will 23
decode the cues the intended way, whereas numerous studies showed interpersonal differences with interpreting some emoji (see chapter 4). In the following work, Walther (1996) proposed a hyperpersonal com- munication theory that tried to solve the question of why people can form online relationships despite less rich media. He claims that the online environment can lead to even closer relationships (hence hyper- personal) than face-to-face communication in some cases. That is be- cause senders can selectively display themselves and show only likable parts of their personalities. Another contributing factor is mostly asyn- chronous nature of CMC, which allows senders to edit their messages more. As a result, more idealized relationships are created that are per- ceived as closer despite less information exchanged. In other words, Walther’s theories (which he primarily supported by experiments) showed that the meanings of CMC messages are less straightforward than face-to-face communication. He later translated this notion into his experiment regarding emoticons (Walther & D’addario 2001), which concluded that while :-) emoticon increases the perceived positivity of a positive message, the frowny emoticon :-( does not have a directly opposing effect. While it (as expected) lessens the positivity of a textually positive message, it does not increase the nega- tivity of a negative message. This result suggests that emotions project- ed into emoticons do not have the precise same effect as corresponding human expressions. However, media and communication theories alone are not sufficient in emoji and emoticon research as the usage also touches different individual and cultural aspects. 3.2 Social and psychological theories According to Habermas (1984), as a goal-driven behavior, human com- munication can serve to understand two people better or create and maintain relationships. Considering emojis and emoticons as a part of communication processes, Tang and Hew (2019) identified several the- ories and models in other social sciences applied in previous emoji and emoticon research. 24
Firstly, uncertainty reduction theory (Berger & Bradac 1982) is often used in studies focusing on the understanding and meaning of messag- es, including emoticons/emojis in interpersonal communication. It shows that CMC users actively seek cues about a person from their messages and predict their behaviors and attitudes – emojis and emoti- cons are such cues, too. When applied to email communication, Byron and Baldridge (2007) found that senders are perceived more likable when using emoticons. Thus, emoticons and emojis add an essential level of information to the message that the receiver uses to form an impression of the sender. Secondly, the intimacy model is presented by Tang and Hew (2019). Like Byron’s and Baldridge’s (2007) findings, several studies linked higher intimacy levels to the usage of emojis and emoticons. Research has shown that when one describes their emotions in any way, it strongly impacts perceived intimacy (Laurenceau et al., 1998). As emot- icons and emojis are also used to show emotions, as another study proved (Janssen et al., 2014), they alone also impact intimacy between CMC users. Hence, emoticons and emojis can function both as a tool leading to forming closer relationships and as an external expression that the relationship between the communicators is close. Thirdly, in line with Walther’s and D’addario’s research (2001), which proved the disparity between expressing positive and negative emo- tions, others linked politeness theory (Brown & Leviso, 1987) with emoticon research. The idea says every person wants to show the bet- ter side of their personality and “avoid threatening another’s person’s face” (Tang & Hew 2014). As one study proved (Skovholt et al., 2014), senders use emoticons to lighten the tone of requests or directives. Importantly to this thesis, Tang and Hew (2014) also mentioned previ- ous researchers exploring how people used emojis differently across cultures and pointed out that more “culturally situated research is needed” on emojis, emoticons, and stickers. The following chapters will show that such research has been conducted since 2014. As emoticons and emojis were subject to cultural differences from the beginning, they can also be parts of social norms, either in smaller groups or on a national and cultural level. 25
3.3 Cross-cultural theories As chapter 2.2. addressed, emoticons were from the very beginning subject to cultural differences due to different designs. This fact alone poses a foundation for further analysis of how emoticons and emojis are used in other parts of the world, no matter the influence of individ- ual and psychological factors of senders and receivers. This chapter seeks to introduce the main concepts on the intersection of cultural and emoticon/emoji studies. Cultural studies have roots in social sciences, political economy, or phi- losophy. Thinkers from various fields noticed that people in different regions have different thinking, working, and communicating styles. For example, psychologist Richard E. Nisbett (2004) concluded that Asians and westerners “have maintained very different systems of thought for thousands of years” (p. 16). Others have tried to quantify and measure these differences and devel- oped models that compare people’s behavior in different cultures. One of the first attempts in these respects was conducted by anthropologist and cross-cultural researcher Edward T. Hall3. In his work, he divided cultures into a continuum between high-context (HC) and low-context (LC) based on how implicitly or explicitly information is transmitted between senders. In other words, he examined if every necessary in- formation is included in an exchange (LC), or instead its members count with an individual having some prior context (HC). Some typical HC cultures are China or Korea; LC culture is the USA or Germany. Both verbal and non-verbal communication in HC cultures typically in- cludes a lot of hidden meanings and subtle signs that are not under- standable for people outside such cultures. These cultures, therefore, tend to be more homogenous. On the other hand, communication in LC cultures refrains from subtle meanings or hidden signs and tends to be explicit and direct. In other words, such communication is usually un- 3 (1914-2009). Do not confuse with a sociologist Stuart Hall (1932-2014) 26
derstandable even for outsiders and can be observed in more hetero- genous cultures, cities, or countries (Hall, 1960, 1989). This concept has been used in numerous fields, even in CMC: for exam- ple, Elizabeth Würtz (2005) performed a cross-cultural analysis of web- sites from HC and LC cultures using Hall’s model, and Park and El Mimouni (2020) applied it to study of emoticons in Arabic, English, and Korean tweets. It is not the only dimension that Hall has used to evalu- ate culture. He also observed a.) differences in proxemics, i.e., studies of how closely people are standing to each other in personal or business encounters and b.) differences in perception of time: in some cultures, people work with relaxed schedules, whereas in some, they are very punctual. In another attempt, Dutch social psychologist Geert Hofstede construct- ed a more complex theory of cultural dimensions. The most recent ver- sion of the model (Hofstede, 2010) contains six dimensions of culture evaluation: 1.) power distance (meaning the higher it is, the more estab- lished hierarchy; lower means people are questioning the authority), 2.) individualism vs. collectivism (how much integrated are groups in each culture, how much is “I” or “We” emphasized), 3.) uncertainty avoidance (how much are people prone to new, unsure things; the higher the score, the more they rely on absolute truths, laws, guidelines), 4.) mas- culinity vs. femininity (masculinity as preference to achievement, hero- ism, success; femininity as preference of cooperation, modesty, caring, quality of life), 5) short-term vs. long-term orientation (short: orienta- tion to past, traditions, personal steadiness; long: orientation to future, adaptation, investment) and 6.) indulgence vs. restraint (indulgence: orientation to freedom, human desires, enjoying life, positivity; re- straint: control, regulation, strict social norms) (Hofstede, 2011). Hofstede firstly developed the theory as part of his research of values of 117,000 employees of IBM company in the 1960s and 1970s, from 66 countries, but later studies in numerous settings followed (for example, Hofstede 1993, 1994; Minkov, 2007; Mouritzen & Svara, 2002). However, as widely as Hofstede’s model has been used, it has also been substantially criticized. The most cited critique is made by Brendan McSweeney (2002). He points out the number of methodological flaws 27
present since the original research in IBM. For example, there were fewer than 50 participants in some of the countries. McSweeney also criticized that Hofstede endorses strong national cultural determinism and implies that culture is universally shared in a nation or rather a state. Also, the very fact that Hofstede works only with nations and no other cultural units is criticized. For example, the model sees Great Britain as a unified culture, even though it consists of four unique coun- tries. Hofstede replied to the critique (2002). While he acknowledged nations are not the best units to study cultures, he claims they are “usually the only kind of units available for comparison and (are) better than noth- ing.” He also defended the use of subsidiaries of one company (IBM) as proxies for national cultures. He argued that the similarity of groups between cultures is essential and argued that the model was validated even in the case of data obtained from representative samples of entire national populations. Although criticized, though, Hofstede’s model is still used. To the best of my knowledge, contemporary cross-cultural studies do not have similar and widely cited frameworks to evaluate cultural differences. The most significant limitation of the model seems to be that it has been devel- oped in the work environment and tries to describe fundamental cul- tural differences. The very same approach was repeated later by Erin Meyer (2014), who has a background in business studies as well and came up with eight dimensions that incorporate both parts of Hof- stede’s and Hall’s models. Moreover, Hofstede’s model is also used widely in CMC. Related to the topics of this thesis, a group of Chinese researchers (Lu et al., 2016) used Hofstede’s model to compare emoji usage in 212 countries. They found, for example, that users from strong-individualism index coun- tries are more likely to express positive emotions through positive emojis than negativity through negative emojis, or that “users from high uncertainty-avoidance countries are less likely to express positive emo- tion” (p. 778). Especially the correlation between the emoji usage and position of a given country on an individual-collectivism continuum can be expected 28
in large. Nisbet and Masude (2003) pointed out that people in individu- alistic and collectivistic cultures differ in communication. In individual- istic countries, individuals are usually pushed to express their feelings explicitly. On the other hand, individual communication is more dis- creet and indirect in collectivistic cultures. Therefore, one can expect more emojis and emoticons used in individualistic countries. In general, emoji and emoticon research have a strong foundation in media and communication theories, and its effects can be evaluated by several theories from other social sciences. Research in the context of cross-cultural studies is also present but inherently limited by the rela- tive underdevelopment of the field in the context of CMC. The following systematic review will also examine if researchers use any of the pre- sented cultural models to explain differences in emoji and emoticon usage. 29
4 Emoticon and emoji as a research subject Emojis and emoticons have raised substantial interest among research- ers and are vibrant research topics. As of September 2021, according to Web of Science (WoS), over 2,120 academic works contained emojis or emoticons in their title (Web of Science, 2021). Google Scholar search shows 454 papers with emoji or emoticon in their abstract or title, just in the past year (Google Scholar, 2021). In general, most papers on emojis and emoticons are from the 2010s or later, even though emoti- cons have been around for decades. This may also explain why research on emoticons seems to be relatively underdeveloped, as opposed to emojis. Some of the previous research on emojis and emoticons is complicated by terminology issues, as Tang and Hew (2019) pointed out. There is widespread confusion between the terms emoticon and emoji. Some scholars use emoticons for all types of nonverbal cues, including emojis, even though they are technically different. This may be caused by the similarity of both words or the fact that some platforms automatically convert emoticons such as :-) to J. Some researchers also combine evaluations of emoticons and emojis. To date, there have been two systematic reviews regarding emoji and emoticon research. Tang and Hew (2019) took a coherent approach and systematically reviewed 51 studies on emoticons, emojis, and stickers (together dubbed as graphicons) conducted between 1996 and 2017, using qualitative analysis. Their conclusions were summarized into three categories: 1.) why do people use these graphicons, 2.) how are they used, 3.) and what is their impact. As part of the first category, they found that all graphicons are used to express emotions, for enjoyment, fun, and social purposes, to substitute textual expression, but also to avoid misunderstanding. In line with that notion, the study also con- cluded that people tend to use understandable emoticons in positive contexts. It also found that females tend to use more emoticons in fre- quency, but males use a more expansive range of them. In general, the study also found out that, while graphicons impact how the message or the sender’s personality is perceived, in situations when the text con- flicts with graphicons, the nonverbal cues do not override the meaning 30
of the text itself. The authors also pointed out to four literature reviews on emoticons (Aldunate & González-Ibáñez, 2017; Derks et al., 2008, Dunlap et al., 2016; Jibril & Abdullah, 2013). None of them, however, were systematic in their searches. The second systematic review was conducted by Bai et al. (2019). They state that “in recent years, emoji have become a hot topic for research, with the volume of papers increasing gradually from 2015 and peaking at 2017-2019.” In total, they included 167 articles from journals and con- ferences, but the review provides only a general overview, without any main synthesis and conclusions. The study explored the development of emojis, their use, and their functions on the personal, platform, and cul- tural levels. 4.1 Cultural predictors of emoji and emoticons Both reviews mentioned above (Bai et al., 2019; Tang & Hew 2019) in- clude parts where studies exploring emojis and emoticons from a cross- cultural perspective are briefly summarized. This fact further proves the relevance of the topic of this thesis. Previous research examined differences in several cultural cohorts: large regions (such as east vs. west), languages, countries, or cultural groups within one country. Spe- cifically, the researchers observed differences in emoticon and emoji types, their frequency of usage, and interpretation. 4.1.1 East vs. West differences The research regarding differences in emoji and emoticon usage be- tween east and west regions comes from the historical notion that emoticons design has separate traditions (see chapter 2.2.). For exam- ple, Park et al. (2014) conducted a global analysis of 1.7 billion tweets from 2006-2009 and focused on how users from individualistic and collectivistic cultures had different preferences (in general, while indi- vidualistic cultures are western, collectivistic are eastern). Using Hof- stede’s cultural model, scholars concluded that people from individual- istic cultures indeed prefer vertical emoticons like :-) and people from collectivistic cultures prefer horizontal ones like ^_^. 31
Ten years after the study by Park et al., another group of researchers (Guntuku et al., 2019) tried to see if a similar logic holds in the case of emojis. They selected online messages from Twitter and Weibo. China and Japan represented the east, while USA, UK, and Canada represented the west. Performing both frequential and semantic analysis, they found out that, in general, usage in both regions was somewhat similar. For example, the order correlation for the most frequent emoji in both are- as was .745. In terms of semantics, it was .59, meaning there is some normativity regardless of the region, though this is still less normativity than correlations between west countries. Tang and Hew (2019) pointed out that there were also several instanc- es when more specific usage patterns were observed, for example, in the case of sequencing emojis. In the west, the sequences primarily re- peat the meaning of the message (for example, “I love you ❤ ❤❤“) to amplify it (Gawne & McCulloch, 2018), users on the Chinese social net- work Weibo are more prone to using sequences with their unique meanings, irrespective of the text around it (Ge & Herring, 2018). For example, in the east, a sequence like means “I do not have beer and ice cream.” Altogether, previous results show significant dif- ferences in usage patterns between east and west, but not too confusing or different interpretations. 4.1.2 Language and country differences More frequently, scholars chose to compare emoji and emoticon usage in two or more languages or countries. Vandergriff (2013) found that nonverbal cues, such as interpunction or emoticons, are context de- pended, thus not usually carrying a meaning on their own. This means they also inherently rely on the text and language they are presented within. More subtle differences were found in the case of research by Barbieri et al. (2016a). Their research studied emoji usage in American English, British English, Italian and Spanish. While overall semantics were preserved between languages, the research found that speakers of Italian interpret emojis slightly differently than the speakers of other languages. Several studies presented comparisons between two countries. For ex- ample, Roele et al. (2020) compared usage in Netherlands and England 32
and found the user patterns and interpretations in both countries are similar. 4.1.3 Intra-country and subculture differences Some researchers tried to examine differences in usage patterns within one country or a subcultural group. Barbieri et al. (2016b) compared tweets from Barcelona and Madrid but found no semantic differences between the two cities. In the case of lesser-used emojis, there were some subtle differences in usage context. Others investigated groups in eastern countries. For example, Zhou et al. (2017) found in a qualitative study that compared users from urban, rural, and small-town areas in China. They discovered that emojis and stickers convey important cultural meanings specific to those areas. China is, after all, distinct in its emoji and emoticon usage, even due to its political climate – some journalistic reports explained, for example, how emojis are used to avoid censorship (Andersen, 2018). In general, cultural differences in emoji and emoticon usage can be the- oretically observed on all levels of cultural differences, whether be- tween fundamental parts of the world, individual countries, regions, or languages. The following systematic review will therefore explore all such levels. Emoticons are culturally dependent by nature, and in the cases of emojis, their understanding seems to be globally rather univer- sal. However, in both cases, some differences can be expected in the number, frequency, and grouping. Such disparities could be attributed to the people in different cultures communicating differently, for exam- ple, more implicitly or explicitly (in line with Hall’s model) or openly or discreetly, in individual and collectivistic cultures from Hofstede’s model. 4.2 Individual predictors of emoji and emoticons Although my thesis is not concerned with predictors of usage other than large group (culture, nation) identity, it is also important to briefly address the extent to which the usage patterns differ due to different individual characteristics, as they may interfere with the cultural pre- 33
dictors themselves, and they are frequently examined in emoticons and emoji research. Bai et al. (2019) state that many more factors, including gender, influence the personal use of emojis. One study found that while males and females understand the functions of emoji the same way (Herring & Dainas, 2018), females use emoji more frequently and in a positive context (Prada et al., 2018). Women also tend to perceive emo- jis as more familiar, precise, and meaningful (Rodrigues et al., 2017). Also, Butterworth et al. (2019) found that interpretations of emoji vary depending on the gender of the sender: “Texts with affectionate emojis were judged as more appropriate and likable when they came from women than from men” (p. 1). These results suggest that emoji usage can be differentiated not only by cultures but also by other sociodemo- graphic categories. Bai et al. (2019) also point out other contributing factors that compli- cate using emojis for communication. There are, of course, several psy- chological aspects related to, for example, personality traits or extra- version (Hall & Pennington, 2013). The perceived valence of an emoji is also a problem: Berengueres and Castro (2017), for example, pointed out that the same negative emoji may feel different in terms of negativi- ty between sender and receiver. Confusion between senders and re- ceivers can also be present because of different expectations. Thomson et al. (2018), for example, focused on the meanings that emojis like , , or have in sexual contexts, even though their original pur- pose was different. The research found that in the case of some emojis, up to 34 % of people receive them in a sexual context. Such results im- ply that an emoji’s correct interpretation may rely also on collective, culturally created meaning. Other factors include specifics of a platform. For instance, Miller (2016) explored how users on different platforms interpret the “grimacing face” emoji differently because of the non-consistent ways media ren- der this character. On some platforms, it was perceived more positively, while some people find the character negative. Also, users on different platforms tend to have their unique preferences in the usage of emoji (Tauch & Kanjo 2016). 34
4.3 Emoji as an objective research tool Not all research that includes emojis counts with usage disparities be- tween individuals and bigger groups. Some researchers assume that interpretations and emotional valences of emojis are universally shared and use them as an objective research tool to measure other values across countries. For these purposes, some scholars annotated emojis according to their perceived negativity or positivity. For instance, Krajl Novak et al. (2015) classified 751 most common emoji by using human annotation and, Kimura and Katsurai (2017) assigned multidimensional emotional vectors to emoji, in a similar manner as previously explored by Aoki and Uchida (2011). Also, Mayank et al. (2016) arbitrarily divid- ed emoji into clusters representing different human emotions. Such models that determine each’s emojis’ emotional valence can be used in many ways, including those having little to do with the emojis themselves. Emojis as objective tools have also been used in fields clos- er to the topic of this thesis; such are sentiment analysis, performed on CMC texts that include emojis (Sari et al. 2014; Felbo et al., 2017). Some researchers assume their independence on language, so they used them for cross-language validity (Guthier et al., 2017). Al-Azani (2018), for example, used emoji to analyze sentiments of Arabic tweets, hence also pre-giving emojis emotional values in the process. This subset of re- search is by no means unsatisfactory, but it inherently assumes that emojis are perceived and therefore used the same way by everybody. Because of that, the following systematic review will not include those studies. 4.4 Research question Emoji and emoticon research are widespread and touch many different fields. Preliminary literature search suggests that some regions, such as China, may be already more investigated, while other countries can ul- timately be left behind. As the previous chapter also showed, there are several levels of cultural differences: between regions, languages, coun- tries, or even cities. All of them can function as distinctive cultural units. However, they are not exclusive, and the effects of each may overlap, and the following systematic review will include all of them. The term 35
cultures in the following research question incorporate all these cultur- al units. The ways the differences between cultures are observed also differ and can also be conceptualized in a number of different ways: previous studies measured frequency of emojis or emoticons as a percentage of digital text, diversity of the most used, rankings of the most used ones, perceived emotional valence, or interpretations by members of differ- ent cultures. These measuring methods are summarized under patterns of usage in the following research question. The research question is therefore as follows: RQ1: What differences in patterns of emoticons and emojis usage exist between cultures? As the RQ suggests, the aim of the following systematic review will be mainly exploration and summarization, as no such attempt in this sub- field has been made. Also, the systematic review will explore whether researchers use some of the presented cultural models. 36
5 Method The systematic review was performed based on PRISMA guidelines that ensure replicability and transparency of the research (PRISMA, 2020). First, to find studies that would be included in the review, the keywords were defined based on papers mentioned in previous parts of this the- sis: emoji, emoticon, culture, cross-culture, country, language, region differ- ences In order to gain the most relevant results, the keywords were con- structed into a search term using Booleans. Four databases were com- bined: Web of Science, Scopus, Proquest, and Ebsco Finder. Because of the specific user interfaces of the databases, the search terms were slightly altered to each database to search for similar studies effective- ly: • Web of Science and Ebsco Finder: (emoji* OR emoticon*) AND (cross-cultur* OR cross cultur* OR cul- tur* diff* OR cultur* OR language diff* OR country diff* OR countr* OR region) • ProQuest: [STRICT] noft((emoji* or emoticon*) and ("cross- cultur*" or "cross cultur*" or "cultur* diff*" or cultur* or "lan- guage diff*" or "country diff*" or countr* or region)) • Scopus: TITLE-ABS-KEY ( ( emoji* OR emoticon* ) AND ( "cross- cultur*" OR "cross cultur*" OR "cultur* diff*" OR cultur* OR "lan- guage diff*" OR "country diff*" OR countr* OR region ) ) AND ( LIMIT-TO ( DOCTYPE , "ar" ) OR LIMIT-TO ( DOCTYPE , "cp" ) ) The four searches were performed on 10th November 2021, on full texts of the studies. An exemption has been made in the case of the Scopus database, which firstly produced over 2000 mostly irrelevant studies. Hence, the search was limited in this case to include only mentions in title, abstract, or keywords. The primary inclusion criteria were selected as follows: 37
• the study must be published in a peer-reviewed journal or as a conference proceeding • the study must be in English • full text of the study must be available • the study must focus on the usage of emojis/emoticons in the form of either their frequency/ their popularity or order • the study must provide information on the usage of emo- ji/emoticons in a form allowing comparisons between at least two cultural units While the proceedings are not typically subjected to systematic reviews in social sciences, they are more frequently used in computer sciences. Given the interdisciplinary nature of this thesis, the decision was also made to include them. Chapters from books, diploma theses, commen- taries, and other forms of content were excluded a priori. The systemat- ic review also consists of both objective measurements and self-reports of emoticons and emoji usage. Web of Science 423 Scopus (titles, abstracts, keywords) 191 Proquest 121 Ebsco Finder 135 Total 850 Table 1: Break down of numbers of records found by each database In total, n = 850 studies were found (Table 1 breaks down this number for each database). After removing duplicates, n = 586 studies were identified. In the next step, titles of the studies were reviewed, and those which did not involve CMC, digital communication, or cultural units were removed. The rest (n = 268) was evaluated based on ab- stracts. The irrelevant studies were further excluded if they met at least one of the following exclusion criteria: • the focus of the paper was CMC or communication or cultural studies, but no survey of emojis or emoticons was conducted • emojis or emoticons were the main subjects of the given paper but without any comparison between two or more cultural units (nations, age, etc.) 38
• the article was a sentiment analysis where emojis and emoticons were used as determinants of the emotional valence of a text in the CMC, or emojis were used as anchors in answering scale for questionnaire items instead of other anchors (numbers or word descriptions) (see chapter 4.3.) • the only focus of the paper was on suggesting new emojis in- stead of studying the present ones • results of the paper did not include any findings regarding cross- cultural differences, even though the data itself was gathered one two or more cultural groups Figure 2: Flow diagram of the systematic review 39
Full texts (n = 32) were read in the following step. One paper was omit- ted from the review due to its unavailability in full text, one because it was in Turkish only (abstract was in English). Others were ignored if they met one of the previously stated exclusion criteria. In the end, n = 17 studies were found to be eligible. Two more studies were found in a subsequent search of the citations of these studies (snowball method). Also, an exemption has been made to include one widely cited but not academically peer-reviewed study. SwiftKey (2015) developers probably gathered the first global emoji dataset, which points out some cultural differences. Therefore, the following systematic review will be performed on n = 20 studies. Figure 2 shows a flow diagram that illustrates the selection. Given the exploratory nature of the research question, the results will be synthesized narratively. The main research table (Appendix A, also available as separate Excel file in the thesis archive), which incorpo- rates the systematic review results, consists of several categories that will be observed. Firstly, it will categorize the studies based on the fo- cus on emoticons or emojis. Secondly, the source of the observed emojis will be presented and their processing method. Cultural units that the given study uses will be offered. Notably, the following category will observe if the researchers used any proposed cultural indexes. Then, four specific differences in patterns will be observed on a four- point scale: yes, mostly yes, mostly no, no. “Yes” or “no” mean the study showed clear and consistent results for the respective usage pattern, with no contradictions. “Mostly yes” or “mostly no” mean that the over- all majority of crucial findings were in one direction, but some (minori- ty) contradicting or inconclusive results were also presented. If the giv- en metric was not studied, the “n/a” mark was used. Apart from this coding mechanism, the key findings are summarized in the primary research table (Appendix A) and the Results section. The first pattern is differences in the type of emoticons used by the cul- tural unit, which will prove or debunk the notion of distinct styles of emoticons, for example, horizontal and vertical, used in different cul- tural units. In the cases of research that only study emojis, this category is omitted since there is only one emoji style in the Unicode. 40
Second, total frequency difference by the cultural unit will signify wheth- er the given study found that, for example, Italians are using more emo- jis or emoticons than Japanese. To correctly state results in this catego- ry, the given study has to use an equal size of the research sample, for example, the number of tweets for all measured cultural units. Third, rankings of the most used difference by cultural unit signify whether the study observed if there are differences in the top used an individual emoji or emoticon in given cultures. Lastly, semantic differences by the cultural unit will see if there are any differences in terms of interpretation or semantic usage of individual emojis or emoticons between cultures. All of the differences in these four categories will be described between emojis and emoticons separately, as they are technically different, and the patterns may not hold both in the cases of emojis and emoticons. 41
6 Results Several interesting patterns can be observed from the systematic re- view. This chapter will summarize results in the cases of emoticons, emojis, and their combinations with cultural models. 6.1 Emoticons in cultural perspective Eight4 of the twenty papers (7,8,9,13,14,15,16,20)5 in the systematic review considered cultural differences in type of emoticon usage. Five of them studied the differences in the use of western :-) and east- ern (^_^) styles of emoticons. They unanimously agree that there are indeed differences in usage of emoticons between eastern regions such as Japan and South Korea and western regions, such as English- speaking countries. As chapter 2.2. of this thesis stated, the eastern style of emoticons came to life in Japan, and the studies proved they are prevalent in this country and surrounding ones, too. Park et al. (2013) suggested that emoticon usage differs between East Asia and the rest of the world, a statement which was supported a year later by Park et al. (2014), who concluded that South Korea, Vietnam, Japan, and China are the countries with the highest usage of eastern style emoticons. The tradition of eastern style emoticons may be even more deeply grounded in the respective societies, as Kayan et al. (2006) found that East Asians rated emoticons in an instant messaging as significantly more im- portant than North Americans. Four studies also examined the differences in total frequency of emoticon usage as a percentage of digital text, three of which found ap- parent differences between countries. However, data is too sparse to clear conclusion in individual countries. Plug (2011) found that Indians (high-context culture according to Hall) use more emoticons than Ger- many (low-context country), and Kavanagh (2016) concluded Japanese (HC) use more emoticons than Americans (LC), even though he did not 4 Combination of “emoticons” and “emoticons and emojis” categories in column I in the main research table (Appendix A) 5 Numbers corresponding with the numbers in the main research table (Appendix A) 42
You can also read