Literature Review: Impact and Implications of Emojis In Semantic Change and Use

Abstract

The landscape of language is changing through emojis, and thus, by extension the world as we perceive it. From a linguistic relativity point of view, there is a plethora of potential scholarship to endeavor on emojis — though, 1) finding some firmer ground on what emojis are specifically, and 2) how emojis are used in practical terms — would both benefit the field. Importantly, when considering the following questions: How are emojis semantically mapped in relation to words (lexical items) and other sentential cues, i.e., hashtags? What happens when emojis change their meaning(s), context, and use; and what does that mean for the way users of emojis experience or perceive the world around them? On this, I offer a hypothesis that 1) select emojis are undergoing accelerated semantic changes, with specific sociolinguistic stratifications (i.e., changes led by ‘youth’), and 2) that the direction of change and how the semantics of emojis are mapped, has linguistic relativistic properties and that because of this emojis are impacting us on both an individual and societal level. In other words, emojis are shaping our world.

Herein, I examine the literature, particularly with a critical eye on the most recent relevant scholarship, to better understand and illustrate these effects of emojis. From this look at the literature, it is my opinion that my hypothesis shows support. In closing, I offer a holistic view of where further relevant scholarship would be helpful and/or is needed, in effort to build a clearer and more concise picture of the effects of emojis on human experience, and yes, cognition.

1. Introduction, Approach, and Barriers

First, it should be noted that though this is not an entirely comprehensive review of scholarship, nor is it systematic in its entirety, I believe it to be a good start in examining the topic. Here, my approach to scholarship review was via Google Scholar and the Wayne State University libraries database. On this, I examined literature relevant to searches of: ‘semantics of emojis,’ ‘emojis change in meaning,’ ‘emojis and linguistic relativity,’ ‘meanings of emojis,’ ‘systematic review of emojis,’ ‘emoji corpus/corpora,’ ‘age grading in use of emojis,’ and ‘stratifications of emoji use.’ I found a diverse set of available research, though to be honest, there was quite a bit of study that was not directly relevant to the topic herein. Moreover, some of the literature directly made mention on the difficulty in identifying relevant scholarship in the field.

So, first it is helpful to identify some main strides, as the definition, use, context, environment, and semantic values of emojis are not a ubiquitous concept that are agreed upon by scholars. Establishing what these are, in correlation with some of the barriers to the field of research, both will help in illustrating where scholarship is at, which extends to the available literature reviewed herein.

On this, the digital world, and the ways in which humans use, perceive, imagine, relay and rely upon is a vastly large domain. As such, emojis and emoticons are widespread in the realm of social communications and beyond. For example, the use of a ‘sad face with tear,’ 😥, may imply one thing to someone sending a text message, but then it may imply a totally different thing to the receiver — more on this will be presented below.

The landscape of emojis and emoticons is a non-standardized linguistic aspect of language, in that there are no widespread, nor entirely universal, nor truly agreed upon notions of what emojis are, particularly when it comes to their meanings (emojis are not ubiquitous in how they are semantically mapped). There are, however, a few frequently agreed upon notions, 1) they are pictographic and/or ideographic in nature (meaning they visually depict an object, state, or idea, or they symbolically represent some kind of abstract idea), 2) they have some level of linguistic semantic value (though to what extent is not agreed upon), 3) they are utilized primarily (though not exclusively) through digital communications, i.e., social media or mobile device communications. Interestingly, from the literature herein not much more can be widely established — such as syntax, or if they are in fact secondary/auxiliary linguistic elements, i.e., ‘gesture.’

Before moving into a handful of studies, there are just a few more points to touch on, for a bird’s eye view, as well as some limitations or barriers to keep in mind. First, Li et al. (2019) found that emojis are in fact stratified between cultures, speech communities, and on a more macro-level, by nation/country — here we see that the most frequently used emojis can vary in each respective context. Also, as part of this finding, they found that it is widely demographic and sociocultural aspects (i.e., how individualistic or communal specific speech communities may be) that drive such speech community variances in popular emoji use. For example, in the U.S. (which ranks number one in the world of Tweets with emojis) the use of ‘one hundred (percent),’ 💯, is in the top ten most popularly used, but is absent from the top ten emoji rankings of the following nine most prolific Tweeting (with emojis) countries.

Second, Li et al. (2019) revealed another critical barrier in the field, mainly that social media use and the devices and platforms utilized are not universal, i.e., as of 2019 Egypt does not allow use of Twitter within the country, thus a cross-linguistic or cross-cultural comparison would need to be designed in either some other kind of multimodal study, or statistical approach.

Third, though this may turn out to be helpful in some contexts, emojis can have many meanings, polysemous, and usually are at a minimum co-referential in nature. This is not just between different speech communities or cultures, but also within the same speech communities. Essentially, emojis can refer to entirely different things, and be misinterpreted in meaning within speech communities, as well as on an individual level. For example, Shardlow et al. (2022) showed that the use of the ‘red-heart,’ ❤️, we see widespread variance in either ‘a strong positive emotion’ in/of inter/intrasentential context, or ‘object of warm affection,’ or ‘a beloved person,’ or a ‘feeling of sexual desire.’ This notion of semantic misinterpretation of emojis, between users, will be further extrapolated in the next section.

Fourth, the field is so rapidly being studied and understood that studies just five years ago may in fact be outdated — a point corroborated by several references herein, including that of Li et al. (2019) — who just in 2019 stated to being (to their knowledge) the “very first cross-cultural comparison of the use of emoji based on large-scale Twitter data” (Li et al., 2019, p.1761). Interestingly, they admit that their study and approach is scalable, and can/should be extended beyond Twitter, however, they have not done so.

This fourth point, touches on my fifth and final point on notable barriers (though there are more), in that many datasets or assembled corpora of emoji use are extracted from Twitter, as you will see mentioned in studies presented below. Admittedly, Twitter is a good medium in extracting publicly available data with both text and emojis, but it ranks 15th in most popularly used social media platform (Statista Research Department, 2022), globally, and is not a fair or equitable representation of important sociolinguistic stratifications of emoji use by respective speech communities (Pew Research Center, 2021), such as age, race, economics, and platform use/purpose. Yet, all of these stratifications arguably have a direct impact on emoji applications and change direction. Essentially, when utilizing primarily Twitter to gather data and in study of the field, linguists and other scholars are weakening their findings. Thus, to truly study the hypothesis and further questions presented herein there is a dire need to reevaluate approach and design of relevant scholarship.

2. Landscape and Further Context

To better illustrate and give more context on recent study in the field, I offer a few pieces of secondary source scholarship, before examining more on-topic, semantic mapping studies in the section below. Thus, and first, Lu & Wu (2022) take a somewhat conservative stance on the role of emojis, and how their semantics are mapped or expressed interpersonally. Here, they first perceive emojis as not words, or lexical items, but rather that of auxiliary semantic items, or gesture, i.e., when one uses non-verbal cues in speech. Their stance is that though emojis may be useful in expressing meaning, and communicating thought, they have ‘not formed a fixed or independent system,’ whereby they note that without such a system emojis cannot be respected as on the same level as ‘written language.’ However, they do not address the idea that emojis may in fact be at the early onset of semiotic evolutionary change, and movement toward a more universal symbolic language.

As part of understanding the topic, I felt it necessary to mention Alshenqeeti (2016), whereby it is argued that emojis may be a linguistic tool in better, more universal, forms of cross-cultural semantic/pragmatic values of communication (or between speech communities). On this, there is reference on perceived parallels between emojis today, in digital applications, and that of ancient hieroglyphics and/or cuneiform. Though, this ‘socio-semiotic’ study is arguably dated at this point, as it was published in 2016 with even earlier research referenced. In either case, a somewhat anecdotal argument is presented that emojis may play a larger role in the future as a tool in efforts to break down linguistic barriers, via better translation of non-verbal cues via digital communication (adding a ‘🙄’ [rolling eyes face] to the sentence ‘The lakehouse was a perfect getaway’ may help better translate cross-culturally distinct semantic values, such as sarcasm, irony, or satire). Interestingly, though this study may have some pitfalls, it is strengthened by the more recent work of Nzeaka (2021), of whom found striking conclusions on the universal potentials of emojis, cross-linguistically and cross-culturally. Of note, Nzeaka posited that semiotics, i.e., emojis, may just become a global language, as humans are perhaps ‘rallying to the use of a common language by symbols’ — a fascinating conclusion that also posits that emojis are in an extreme position of semantic change and variation, on a global scale.

3. Implications: Neuro-Cognitive Studies

Briefly, I offer a few studies on cognition (brain activity) as it relates to the relationship between how emojis are semantically mapped and perceived, and what this may mean regarding linguistic relativistic properties of emojis. Tang et al. (2021), conducted a study that showed notable EEG theta responses (event-related electroencephalogram power) induced by emoji semantic violations. Though their participant pool was relatively small, their findings were rather striking. For example, when said text does not align with perceived meaning of utilized emojis in correlation, there is brain activity/electrical responses. Their work “suggests a higher working memory load for monitoring errors, difficulty of form recognition and concept retrieval in emoji semantic processing” — essentially emojis that did not correlate with target intrasentential lexical items took longer for participants to process overall sentential semantics, meanings (Tang et al., 2021).

Adding to the EEG study above, Barach et al. (2021) showed some similar findings. As such, they conducted a linguistic survey/study with the use of eye tracking/movements, of a pool of sixty college students (of whom use smartphones), which utilized certain sentences with target words in correlation (or not) with an emoji, placed at the end of the sentence, and then utilized a control sentence without an emoji present. The tracking of eye movements allowed for them to reveal the ‘time course of semantic processing.’ Their findings showed that participants took longer to process and were fixated for longer periods of time when ‘emojified’ text had incongruent pairing between a target word and a respective emoji. For example, there was a notable difference in cognitive response between ‘My tall coffee is just the right temperature ☕’ versus ‘My tall coffee is just the right temperature 🍺’ (coffee emoji versus beer emoji). Though, their work does not review co-referential or polysemous aspects of emojis, as they seem to avoid this query, they utilized less abstract and more literal emoji semantic interpretations, i.e., ‘cookie’ as 🍪. Likewise, sentences with semantically aligned emojis in relation to target words showed 1) faster semantic processing and shorter periods of ‘eye gazing,’ less refixation, and even at some points emoji ‘skipping.’ From this, it seems to say that their work shows promise, as it reveals a certain neuro-cognitive effect of emojis vis-à-vis semantic processing. Though, I would stress this study may be fluid in its findings when accounting for the possibility that some visually ‘less abstract’ emojis are undergoing rapid change, and perhaps are becoming more abstract in meaning, via change direction by that of younger generations (Pew Research Center, 2021).

Last, the work of Weissman & Tanner (2018), in part, “investigated how humans process semantic and pragmatic content of emojis in real time,” whereby they tested participants via three separate ERP studies, to then analyze P-effect results (brain processes/neural responses). The study compared ‘irony’ values between that of words and emojis. Though I won’t go into the full extent of their studies, in one study they elicited responses of participants based on emojis being correlated (matching) to the semantic value of the sentence, or not correlated (mismatched), or in irony, i.e., ‘The cake she made was terrible 🙁’ as a match, ‘The cake she made was terrible 🙂’ as a mismatch, and ‘The cake she made was terrible 😉’ for irony. Interestingly, they found a strong P200 effect of ironic ‘wink’ emojis, which is on par with the same P-effect level as word irony. As such, they posit that there are parallels between cognitive processing of ‘word-and-emoji-induced irony.’ In other words, there are similar neuro-cognitive effects in semantic processing of lexical items (words) and that of emojis, or ‘ideograms.’

4. Implications: Stratification, Effect, and Change Studies

Here I explore some of the observed changes, change direction, and likely sociolinguistic stratifications of users of emojis. Starting first with perhaps the mildest in relation to my topic herein, to then conclude with more direct evidence on my hypothesis, based on the literature, of age-related or age-grouping differences in emoji evaluation and meaning.

First, Riordan (2017), which borders being outdated in my opinion, found via their ‘text-messages’ based study, that both ‘face’ (i.e., 😊) and ‘non-face’ emojis (i.e., ⏰) are utilized for everyday emotion work, and that the use of emojis may preserve and/or enhance social relationships between users. Interestingly, they also found that both kinds of emojis showed some level of semantic value upon sentential meaning. For example, the use of a semantically correlated non-face emoji alongside a negatively charged sentence may soften the strength of negativity, i.e., the difference between ‘The camel spit on my face’ versus ‘The camel 🐪 spit on my face.’ Last, their work identified emojis as more often than not in the position of adding some level of positivity or ‘softening’ to a sentence meaning — and equated a certain aspect of ‘joy’ in combination with the rest of the sentence meaning. Though, their work largely 1) placed a lot of importance on the use of temporally early on ‘face’ emojis/emoticons (first released to the public emojis), 2) seemingly perceived emojis as ‘gesture’ throughout their study and did not provide strong evidence of this being the case, and 3) did not really get into the possibility of emojis being polysemous so far as their semantic mappings are concerned. Regardless, their work is still a good testament to how emojis in everyday use may impact real world user experiences, mood, and interpersonal relationships — as well as how we semantically map, and even categorize, emotively valued symbolic meanings. Thus, there is already grounds for further examination of linguistic relativity via emojis.

Second, Robertson et al. (2021) conducted a longitudinal study, utilizing computational linguistics of six years of Twitter data, to demonstrate both emoji context and use, but importantly, how emojis are shifting in meaning. From their analysis of the data, they posited that just like how words can shift in meaning over time (semantic change), so are emojis shifting. Additionally, they posited that these semantic mappings of emojis are complex, i.e., they may be multi-directional, and importantly, they are likely moving from simple abstract meanings to greater abstraction. For example, ‘💀’ (human skull) is/has moved (largely between 2013 to 2014) from something more literal in meaning (zombie, corpse, bury, undead, murder, deceased, passed on, died, RIP), to now something more abstract (‘LMAO,’ dead from laughter, hilarious, too funny). Here, I say more abstract because this additional semantic element of ‘humor,’ I would argue, is not something that can be extrapolated from the literal or face-value interpretation of a human skull, ‘💀.’ Also, their findings suggest that it is the seemingly simpler pictographic/ideographic emojis (like ‘👈’, as in ‘backhand index left point’) that are the ones more likely to shift and become more abstract in meaning. For example, ‘🐸’ (‘lizard’ or ‘frog’ in 2012–2013) in just the past ten years has shifted to ‘kermit,’ ‘snitch,’ ‘nvm’ (nevermind), ‘lowkey,’ in 2014, to ‘the n-word,’ ‘bitch,’ or ‘hate,’ in 2015, and most recently, as of 2017, more semantically aligned with ‘spilling the tea’ or ‘gossip.’ This wide disparity in the semantics of emojis in short periods of time, at least with the use and context of ‘🐸’ (and in some, but not all, speech communities) shows the possibility of some serious miscommunication and possible negative real-world outcomes, particularly on an interpersonal level, as the difference between ‘frog’ and ‘bitch’ is quite a disparity.

Building on this disparity, when we start to evaluate stratifications on a demographic level, findings become even more complex and variable. For example, Shah & Tewari (2021) conducted a study whereby they mapped the emoji usage among youth and made some arguably strong conclusions. Their college student survey/study (of 250 participants), which was designed, first, via an in-group focus group, was analyzed to better understand what sociolinguistic elements drive emoji usage. Their findings reveal that there are psychological and socio-behavioral implications of emoji use and strong sociolinguistic stratifications. Here, they found evidence of specific emoji use around ‘level of formality,’ gender, generation gap (age group), semantic strengthening/weakening on an interpersonal level (i.e., mood or tone), and even specific platform variance (Facebook versus Whatsapp). From their study findings, they posit that 1) emojis are often aligned with a certain sociological Information-Agreement Theory (involving emotion sharing online), whereby emojis are used with a specific manipulative intent to elicit a specific audience response or outcome, and 2) more anecdotally, that as emojis become more nuanced (abstract) in meaning, so too, can these shifts be studied and exploited by marketing firms and governments in manipulative societal outcomes. Essentially, emojis are powerful, as they transcend words, images, and that of emotion.

Even more fascinating, Weiß et al. (2020) conducted a survey/study that elicited responses from 170 participants (with different age grading) on the meaning representations of face-based emojis. Their findings, vis-à-vis an exploratory ‘factor analysis,’ revealed notable evidence of ‘age related differences in emotion classification’ via use of and perception of arguably emotively valued emojis. For example, the perceived highest/strongest emotional value of ‘😗’ (kissing face) by younger adults (≤38 years old) was roughly semantically mapped to ‘joyful,’ whilst older adults (≥38 years old) mapped this emoji to be more emotively akin to ‘content.’ Though, their study was relatively limited in breadth and pool of participants, they do posit an important possibility — that age-groups and/or age-grading within a speech community may perceive the symbolic representations of emotion, via emojis, with divergent values. Of note, though their finding was not fully conclusive, it seemed to show some additional level of divergence between positive and negative valence of age-groups, whereby positive valence difference(s) was near null, but negative valence of emotion-to-emoji mapping seemed to increase with age. In other words, they argue that on some level, older aged users of emojis may perceive certain face-based emojis, like ‘🤢’ (disgusted face), more negatively than younger users.

5. Conclusions

From the multiple studies and diverse literature reviewed above — I posit that there is strong evidence that emojis are both undergoing accelerated semantic changes, with age-graded stratifications (among other sociolinguistic stratifications) in relation to the direction of change, and that the application, use, and semantic mappings of emojis all show some linguistic relativistic properties. Here, I would make a case that emojis are somewhere in between a linguistic ‘augmenter’ of the human experience/worldview, and that of a linguistic ‘meddler.’ As such, emojis have been shown, from above review, to cause a shift in the way we perceive and experience things, such as Weiß et al. (2020), which showed that emojis allowed for select age-groups to perceive human emotions differently, such as how young study participants across the board saw the emotional semantic mappings of face-based emojis to be less negatively valued compared to older generations. It can also be argued that emojis may act as a bit of a nuisance or linguistic meddler on how semantics of emojis are mapped. As such, Robertson et al (2021) showed that the use of ‘🐸’ could translate within a speech community as a highly negative meaning, like ‘bitch’ or a racial slur, to a more literal visual interpretation as just a ‘frog,’ or even a specific frog, like ‘kermit.’

Additionally, regardless of the level to which we find support of linguistic relativity, there have already been other studies, relevant to the semantic mapping of emojis, in the broader field. Such as that of the work of Hauthal et al. (2019), which found emojis (collected from Twitter data) qualitatively useful in identifying and collecting emotional and attitudinal event-based and/or location-based responses. For example, their work was able to capture the relative emotional and attitudinal positions (i.e., sadness, or fear, or joy) of study participants across the U.K., regarding participant feelings or positions on the occurrence of ‘Brexit,’ and then were able to map these attitudinal findings. An exceptional sociopolitical tool/finding — all from emojified Twitter data.

Another insight, on the development of the field, is that if we collectively take the stance that emojis are primed to be a semiotic global language, and if we agree that there are linguistic relativistic elements to that of emojis (how we may experience or perceive human emotions vis-à-vis symbols and/or emotive-valued ideograms) — then we may also attempt to try and standardize these universal tools. Standardization is not something I am particularly fond of, as it doesn’t really promote diversity of language, and thus by extension diversity of cultures, but there are potential positive outcomes, too. For example, studies like that of Koh et al. (2019), looked at how to develop/improve certain human-to-computer software which would better translate the intimacy and expressiveness (emotionally valued meanings) of VCMs (i.e., emojis/emoticons). Thus, more broadly make user experience(s), interpretations, and access of emojis more universal, global.

Personally, I think there is a ton of directions this field of research can examine. For example, and of interest, Honkanen & Müller (2021) examined the distribution and frequency of select emojis by bi/multilingual participants in Nigeria to better understand how linguistic ‘interjections’ are distributed, as well as to better understand how a particular semantic item, such as ‘surprise’ (along with emojis that are valued as ‘surprise’) may act or be distributed online, in multilingual societies/settings.

6. Further Scholarship

As touched on at the top of this review, I think it would be beneficial for future scholarship to find creative ways in extracting and/or aggregating data and new corpora. Twitter is less relevant today than it was ten years ago. Understandably, there are limitations when we consider that private companies hold immense amounts of data on their users, i.e., Google, Apple, Microsoft, among many specific platforms— and most of this data is not publicly available. Establishing some sort of incentive or agreement for access to these private data could dramatically change how we are analyzing and studying the field. Additionally, it behooves us to consider the likely realities that semantic change direction of emojis is stratified by demographic variables, especially by age-group.

Part of this is also understanding the disparities between what is defined within these available private/public datasets, in comparison to the actual truth or reality in relation to these data. For example, certain social platforms provide age-groups of their users (sometimes starting at 18 years old) for legal or other reasons — yet, this is a misleading picture, as it does not justly represent the actual ages of the users, merely the ‘registered’ or ‘reported’ ages. Scholarship that attempts to circumvent this sociolinguistic disparity between reported versus actual age of users would be helpful. Last, there could simply just be more studies, with larger participant pools, and larger and more trustworthy corpora (include platforms like Tiktok, Youtube, Facebook, and Whatsapp). Some of the studies presented herein, within the last couple years, claim they are the first of their kind — so adding more voices and more studies on these topics could be critical in better understanding the role and reach of emojis within the human language, thought, and experience — because we would be able to compare similar studies.

I’ll conclude that the field of emojis and their meanings, in regard to the extent to which language may meddle or augment the ‘user’ experience, is an incredibly fast-paced linguistic field, because of the medium. Things are instant, reactions are often live, or in real-time. This is not the same situation as someone in ancient Babylon carving some kind of cuneiform into a stone tablet and then disseminating the information on that tablet over a period of (comparatively lengthy) time. Emojis, and social media more broadly, are unique in their placement, and language contact. As such, I think a large development in the field, that would aid in future scholarship is more resources and accreditation given to that of some kind of ‘emojiology’ field, akin to that of etymology, as I stand firmly in the camp that emoji meanings are both polysemous in nature and are currently in an accelerated state of directional change and variation of their diverse applications, uses, and perceived semantic/emotive values.

References

Alshenqeeti, H. (2016). Are Emojis Creating a New or Old Visual Language for New Generations? A Socio-semiotic Study. Advances in Language and Literary Studies. Volume 7(6). 56–69. doi:10.7575/aiac.alls.v.7n.6p.56

Bai, Q., Dan, Q., Mu, Z., & Yang, M. (2019). A Systematic Review of Emoji: Current Research and Future Perspectives. Frontiers in Psychology. Volume 10. doi:10.3389/fpsyg.2019.02221

Hauthal, E., Burghardt, D., & Dunkel, A. (2019). Analyzing and Visualizing Emotional Reactions Expressed by Emojis in Location-Based Social Media. International Journal of Geo-Information. Volume 8(13). doi:10.3390/ijgi8030113

Honkanen, M., & Müller, J. (2021). Interjections and emojis in Nigerian online communication. World Englishes. Volume 40. 611–630. doi:10.1111/weng.12544

Koh, J. I., Cherian, J., Taele, P. Hammond, T. (2019). Developing a Hand Gesture Recognition System for Mapping Symbolic Hand Gestures to Analogous Emojis in Computer-Mediated Communication. ACM Transactions on Interactive Intelligent Systems. Volume 9(1). 1–35. doi:10.1145/3297277

Li, M., Chng, E., Chong, A. Y. L., & See, S. (2019). An empirical analysis of emoji usage on Twitter. Industrial Management and Data Systems. Volume 119(8). doi:10.1108/IMDS-01–2019–0001

Lu, Y., & Wu, J. (2022). An Empirical Study on the Use of Emojis by College Students From the Perspective of Symbolic Interactionism. Theory and Practice in Language Studies. Volume 12(4). 707–714. doi:10.17507/tpls.1204.11

Nzeaka, E. E. (2021). New Media Semiotics and the Rise of Universal Symbolic Language: Digital Natives’ Perspective Amidst Covid-19. Journal of Higher Education Theory and Practice. Volume 21(1). doi:10.33423/jhetp.v21i1.4048

Pew Research Center. (2021). Social Media Fact Sheet. In Pew Research Center [website]. Retrieved from https://www.pewresearch.org/internet/fact-sheet/social-media/

Riordan, M. (2017). Emojis as Tools for Emotion Work: Communicating Affect in Text Messages. Journal of Languages and Social Psychology. Volume 36(5). 549–567. doi:10.1177/0261927X17704238

Robertson, A., Liza, F. F., Nguyen, D., McGillivray, B., & Hale, S. A. (2021). Semantic Journeys: Quantifying Change in Emoji Meaning from 2012–2018. Association for the Advancement of Artificial Intelligence. doi:10.48550/arXiv.2105.00

Shah, R., & Tewari, R. (2021). Mapping Emoji Usage Among Youth. Journal of Creative Communications. Volume 16(1). 113–125. doi:10.1177/0973258620982541

Shardlow, M., Gerber, L. & Nawaz, R. (2022). One emoji, many meanings: A corpus for the prediction and disambiguation of emoji sense. Expert Systems with Applications. Volume 198. doi:10.1016/j.eswa.2022.116862

Statista Research Department. (2022). Most popular social networks worldwide as of January 2022, ranked by number of monthly active users. In Statista [website]. Retrieved from https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-of-users/

Tang, M., Zhao, X., Chen, B., & Zhao, L. (2021). EEG theta responses induced by emoji semantic violations. Scientific Reports. Volume 11. doi:10.1038/s41598–021–89528–2

Weiß, M., Bille, D., Rodrigues, J., Hewig, J. (2020). The value of a real face: Differences between affective faces and emojis in neural processing and their social influence on decision-making. Experimental Aging Research. Volume 46(5). 416–432. doi:10.1080/0361073X.2020.1790087

Weissman, B., & Tanner, D. (2018). A strong wink between verbal and emoji based irony: How the brain processes ironic emojis during language comprehension. PLOS ONE. Volume 13(8). doi:10.1371/journal.pone.0201727