The Semiotics of Emoji is a monograph that attempts to show that emoji are a new visual language. The 10 chapters approach emoji from different linguistic levels. Chapter 1 compares emoji to alphabetic and non-alphabetic writing systems from history. Chapter 2 looks at the different uses of emoji, focusing on emotional, phatic, and cultural functions, while Chapter 3 argues for an emoji competence found primarily among young people. The next three chapters look at linguistic features of emoji, including semantics (Chapter 4), grammar (Chapter 5) and pragmatics (Chapter 6). Chapter 7 provides a discussion of emoji change, while Chapter 8 looks at the spread of emoji. Chapter 9 draws on some rather contrived examples to discuss emoji in the context of universal languages like Esperanto, while Chapter 10 provides a lackluster conclusion that “Every book needs a conclusion. But in this case, it is difficult to find one.” (p. 181).
In an effort to show that emoji are a new visual language, Danesi draws on three primary data sources. The first is a corpus of 323 text messages contributed by 100 undergraduate students (50 male, 50 female) at the University of Toronto, which are quoted both in aggregate and as illustrative examples throughout the book. The second is a series of interviews with the same students about their emoji use and attitudes. The third is various screenshots of text messages which were found by a research team of four students from around the internet and are presented by Danesi – with little or no indication of their origins – as evidence for some of his most interesting claims.
The first and second data collection methods are relatively standard, if somewhat sparse in their execution for a full-length book. They enable Danesi to make several reasonable (though not especially novel) arguments, such as that most emoji accompany words and are interpreted as markers of tone, that the smiley face is the most common emoji, and that books like Alice In Wonderland, when retold in emoji, are more difficult to understand than hybrid word-emoji passages.
However, there remain several shortcomings in terms of data methodology. For the corpus side, it is a pity that Danesi did not take inspiration from the excellent “big data” corpus work that is common in studies of Computer-Mediated Communication, such as Tagliamonte and Denis Reference Tagliamonte and Denis2008, Pak and Paroubek Reference Pak and Paroubek2010, Schnoebelen Reference Schnoebelen2012, Pavalanathan and Eisenstein Reference Pavalanathan and Eisenstein2016, Tagliamonte Reference Tagliamonte2016, and McSweeney Reference McSweeney2018, all of which draw on datasets in the tens of thousands of entries rather than in the mere hundreds. For the interviews, the questionnaires would have benefited from a more focused approach on explaining participants’ own specific communicative practices rather than repeating their general assertions about emoji, taking as a model the insightful works of danah boyd (Reference boyd2014) and Kelly and Watts (Reference Kelly and Watts2015).
An illustrative example of the limitations of Danesi's approach is in his analysis of the following sequence of emoji (p. 124), which is attributed to an undated “Partnership for Drugs Free Kids” campaign aimed at Millennials.
Danesi claims that this sequence “can be reformulated in words as follows: ‘I'm tired of drinking or doing things to fit in (like an ant).’ So, ‘I need to be strong and eat the right things and not to take drugs.’ […] The text is indeed interpretable, as we found out with our informants, who easily read it and derived the underlying message from it instantly. […] One informant put it as follows: ‘I know my younger brother would understand this’ ”.
As a fluent emoji user, I questioned the claim that the text was “easily” read and so endeavoured to verify it with my own group of informants, which comprised over a hundred emoji enthusiasts from the Millennial, Gen X, and Gen Z demographics. After many interpretations involving suspicion, ants, working out, and sexual innuendo, several participants collaboratively arrived at the realization that the emoji sequence was a rebus: I (eye) want (one + ant) to (two) fit (bicep curl) in (in-tray) but (peach emoji represents a butt, frequently used in sexual contexts). I (eye) don't (negation) want (one + ant) to (two) smoke (cigarette). This rebus interpretation satisfied the remainder of the participants, who unanimously rejected Danesi's gloss, but nonetheless repeatedly and emphatically pointed out that rebus use of emoji is highly atypical, being most atypical of the youngest groups but somewhat more common among the older ones. Indeed, the participant who finally solved the last piece of the puzzle reported drawing on her knowledge of cryptic crosswords rather than emoji.
It is with this caveat in mind that we should consider Danesi's third source of data, that of unsourced or barely-attributed images of emoji from the internet. One illustrative example is on pp. 101–102, where Danesi analyzes a screenshot which he claims represents a conversational text message exchange and which he attributes only to “a public domain website” (perhaps a confusion between websites being public and having domain names with the legal concept of public domain). In any case, the reader will doubtless be aware that all major style guides include instructions on citing both websites and public-domain works. It is troubling that Danesi seems to lack this awareness.
The screenshot in question is a single multi-line text message containing emoji in the shape of a businessman with the words “Honey I'm home”, then emoji in the shape of a woman with the words “Welcome home dear”, then a baby emoji with “DA DA”. The reply text message is cut off, but most likely reads “Lmfao” — thus presumably a text message playing with emoji as a medium and interpreted by its recipient as playful. However, Danesi analyzes it as two genuine text messages sent between a husband and a wife, where “she signs it, not with her name, but with a baby smiley that represents, obviously, their child” (p. 102) and uses it to argue for a conative dimension to salutation emoji.
Similar “stunt” emoji screenshots, attributed to either “public domain websites” or viral content aggregation websites like smosh.com and freemake.com, are found on pp. 36, 38, and 143, and are similarly presented as examples of normal emoji use, despite Danesi's own acknowledgement that no such examples are found in his 323-message corpus. It is for this reason that Danesi would have been wise to construct a larger corpus in the model of Tagliamonte and others cited above, as this might have provided a stronger signal that such examples are indeed atypical. It would also have been wise to curb the urge to cite sensationalist examples from clickbait humour websites as if they were genuine practice – a lamentable tendency in internet language discourse which has been cogently criticized by Thurlow (Reference Thurlow2006) and Schnoebelen (Reference Schnoebelen2018).
Overall, what The Semiotics of Emoji misses is the fact that emoji are interesting not because they are general-purpose small pictures; small pictures have been around for much of human history and many ideas for their use have been proposed to varying degrees of success (e.g., Blissymbolics, Pitman shorthand, clip art, rebuses, and illuminated manuscript marginalia). What makes emoji interesting is that they are a specific set of pictures and that millions of people use them on a regular basis. Danesi is free to write a book about the linguistic potential of rebuses or pictures in general; however, his decision to write a book that claims to be about emoji means that his primary account must be of emoji in their authentic use, much as the writer of a grammar of English cannot use as primary reference texts “Jabberwocky” and Finnegan's Wake.
Regrettably, I cannot end the review here, as I must also call the reader's attention to the fact that The Semiotics of Emoji is riddled with elementary errors. A non-exhaustive sampling:
• “Another problem, albeit somewhat minor, is that emoji characters vary slightly among platforms.” (p. 27, no source provided). By contrast, an experimental study by Miller et al. (Reference Miller, Thebault-Spieker, Chang, Johnson, Terveen and Hecht2016) found cross-platform emoji variation to be a considerable problem.
• The nail polish emoji is discussed on pp. 31–32 without noting its most common nonliteral meaning from African American English (i.e. throwing shade).
• Emoji are presented as a possible cure for dyslexia, with the claim that dyslexia is unknown in China (p. 91). A cursory Google Scholar search for “dyslexia Chinese” finds hundreds of papers to the contrary.
• The Unicode Consortium is erroneously described as a wiki (p. 118).
• A wiki, moreover, is erroneously described as “any website that provides information of a specific kind” (p. 135).
• Three screenshots from animated cartoon series or anime are claimed to be examples of emoji (p. 123), although such full-background images are never found in the Unicode emoji list nor in emoji-inspired sticker apps like Bitmoji. It seems likely that their true nature is most probably animated gifs, but in print reproduction it is of course impossible to tell.
• An image is reproduced as an example of a medieval illuminated manuscript (p. 120) with no citation other than collectorsweekly.com. Investigation revealed that it was posted as part of a list at < https://www.collectorsweekly.com/articles/naughty-nuns-flatulent-monks-and-other-surprises-of-sacred-medieval-manuscripts/>, and is there appropriately attributed to the Rutland Psalter, c. 1260. (British Library Royal MS 62925, f. 87v.).
• Crystal (Reference Crystal2006) is cited as the coiner of the term “netlingo” (p. 154); the term Crystal in fact coined was “netspeak”.
As I am not a semiotician, I confine my assessment to the portions about Computer-Mediated Communication and do not attempt to evaluate the semiotics aspect of this book. However, statements like “Semiotics is used here as a generic tool for evaluating the data, not a technical one” (p. 16), lead one to suspect that practising semioticians may recognize little of their field in The Semiotics of Emoji.
Unfortunately, I can recommend this book neither to scholarly audiences nor to lay audiences. For scholarly audiences, shoddy citation practices and lack of familiarity with research on Computer-Mediated Communication will contribute neither rigour nor insight. For lay audiences, elementary errors about emoji and the internet will make this book frustrate more than it informs.