Emojiphy: The Neo-Hieroglyphy and Emoji Use in 2021

Mind to Market Blog

By Jeffrey Watumull

“One reason for studying language—and for me personally the most compelling reason—is that it is tempting to regard language… as ‘a mirror of mind.’ By studying language we may discover abstract principles that govern its structure and use, principles that are universal by biological necessity and not mere historical accident, that derive from mental characteristics of the species… By studying the properties of natural languages, their structure, organization, and use, we may hope to gain some understanding of the specific characteristics of human intelligence. We may hope to learn something about human nature; something significant, if it is true that human cognitive capacity is the truly distinctive and most remarkable characteristic of the species”

Noam Chomsky, Reflections on Language.

The Unicode Consortium, caretakers of the of the global standard emoji set, have released the most-used emoji of 2021. Each year, the Unicode Emoji Subcommittee parses usage data from around the world to determine What are the most frequently used emoji? What do they have in common? And so, we asked ourselves: in the modern technological age, what can we learn about human nature by studying new forms of language use? In particular, what can we learn from emoji?

Firstly, we learn that its plural form is disputed: emojis (regular inflection) or emoji (irregular inflection)? Empirically, both can be observed. I prefer the irregular on aesthetic and scientific grounds—as does the all-powerful Unicode Emoji Subcommittee. Aesthetically, it “tastes better” on my mental palate; if it and the regular form were for sale in the word market of Dictionopolis —the city of words in the unutterably brilliant novel, The Phantom Tollbooth—it would surely sell like hotcakes, or like “fresh-picked ifs, ands, and buts” and “nice ripe wheres and whens”.

Scientifically, the relative rarity of irregular forms in English increases their interestingness value; and like the Baker Street irregulars in the Sherlock Holmes stories, these oddballs are “sharp as needles” (A Study in Scarlet)—that is, “clever” or effective in how they canvass their metropolises, London and Dictionopolis, respectively. Finally, plural emoji patterns with plural Samurai (e.g., Seven Samurai) and plural Jedi (e.g., “The Jedi are extinct, their fire has gone out of the Universe” (A New Hope)). I rest my case.

The regular/irregular distinction is “self-similar”—in the jargon of fractal geometry—to the general linguistic distinction betwixt combinatorial rules and the memorized representations to which they apply:

“The vast expressive power of language is made possible by two principles: the arbitrary sound-meaning pairing underlying words, and the discrete combinatorial system underlying grammar. These principles implicate distinct cognitive mechanisms: associative memory and symbol-manipulating rules. The distinction may be seen in the difference between regular inflection (e.g., walk-walked), which is productive and open-ended and hence implicates a rule, and irregular inflection (e.g., come-came), which is idiosyncratic and closed and hence implicates individually memorized words.”

Steven Pinker, “Words and Rules”.

These two principles are inherent to our nature:

“A normal child acquires [this] knowledge [of words and rules] on relatively slight exposure and without specific training. He can then quite effortlessly make use of an intricate structure of specific rules and guiding principles to convey his thoughts and feelings to others, arousing in them novel ideas and subtle perceptions and judgments”

Chomsky, Reflections on Language.

The latter capacity “to convey … thoughts and feelings to others, arousing in them novel ideas and subtle perceptions and judgments” has long astonished our greatest philosophers. In his Dialogue on the Great World Systems, Galileo wrote in awe of that capacity to communicate our “most secret thoughts to any other person … with no greater difficulty than the various collocations of twenty-four little characters upon a paper”—an “invention” he adjudged comparable with the creations of great artists like Michelangelo, Raphael, or Titian.

The Port-Royal grammarians agreed in their Grammaire générale et raisonnée: “this marvelous invention of composing from 25 or 30 sounds an infinite variety of words, which although not having any resemblance in themselves to that which passes through our minds, nevertheless do not fail to reveal all of the secrets of the mind, and to make intelligible to others who cannot penetrate into the mind all that we conceive and all of the diverse movements of our souls”.

Returning to emoji, we may conject that by the art of emojiphy—my neologism on analogy to hieroglyphy—we may contemplate its variations:

“The pandemic has affected nearly all aspects of modern life, from the clothes we wear to the food we eat to how we spend our time. There is one thing, however, that has remained almost unchanged: the emojis we send.

The fact that most of the rest of the top 10 in Unicode’s data set, which covers multiple platforms and apps, stayed fairly consistent also signifies just how flexible the current set of emojis are. ‘It basically indicates that we have what we need to communicate a broad range of expression, or even very specific concepts,’ […]. ‘You don’t necessarily need a Covid emoji or a vaccination emoji because you have biceps, syringe, Band-Aid, which conveys semantically the same thing.’”

The New York Times

Of course, the flexibility of emoji to construct “frillions of legitimate news ideas” is inherited from our human language faculty:

It is telling (pun intended) that the flexibility of emoji to represent the new idea of Covid echoes (pun intended) the evolved flexibility of hieroglyphs to represent new ideas not by divining new symbols but by combining existing symbols in novel ways. A beautiful example of this is the oldest known instance of a full sentence written in mature hieroglyphs, discovered on a seal impression in the tomb of Seth-Peribsen at Ulmm-el-Qa’ab, which dates from the Second Dynasty (twenty-eighth or twenty-seventh century BC).

Beyond history, we can excavate our critical imaginations for evidence of how intelligent entities (on Earth or anywhere in the multiverse) have or could communicate complex ideas. The short story, “Story of Your Life”, on which the film Arrival is based, imagines an extraterrestrial intelligence (ETI) with a written language that combines logograms (their words/phrases) not in conformity to any linear logic, but in accordance with principles analogous (or identical) to Fermat’s Principle of Least Time, thus explaining why they do not write a sentence one logogram at a time, but instead draw all of them all simultaneously, implying that they know the complete sentence before the writing of it even begins. If we assume that language determines thought, we can understand that the ETI —and other intelligences that acquire their language—do not experience events in terms of causality and sequentially, but instead experience all events simultaneously in terms of teleology. (When you read the text, be mindful of the curious use of tense by the human narrator fluent in the ETI language.) In the short-story/film, the linguist discerns that each alien logogram is to be divided into 12 sections, and that each subsection enters into what is either a singular expression or sentences of varying complexity. It will be fascinating to see whether human emoji evolve such sophistication.

Returning to terran language, tThe explicitly symbolic form of written human language (e.g., hieroglyphs) is representative (quite literally) of an “enabling function” of written language: by externalizing mind-internal computations, it “extends” the human mind, into the world, offloading the cognitive load so as to optimize the efficiency of cognition. Indeed, philosophers of mind have conjectured radical theories based on this “general tendency of human reasoners to lean heavily on environmental supports. Thus consider the use of pen and paper to perform long multiplication […], the use of physical re-arrangements of letter tiles to prompt word recall in Scrabble […], the use of instruments such as the nautical slide rule […], and the general paraphernalia of language, books, diagrams, and culture. In all these cases the individual brain performs some operations, while others are delegated to manipulations of external media”. The distributed computing of language across brain and environment constitutes an “epistemic action”, which “alter[s] the world so as to aid and augment cognitive processes such as recognition and search”.

I concur with the conjecture of these philosophers—and would extend the conjecture to cover emoji—that epistemic action “demands spread of epistemic credit. If, as we confront some task, a part of the world functions as a process which, were it done in the head, we would have no hesitation in recognizing as part of the cognitive process, then that part of the world is (so we claim) part of the cognitive process” (Clark and Chalmers, “The Extended Mind”). This is conjecture is regarded as radical, although it should not be, for it cannot be doubted, as Darwin explained in The Descent of Man, that language used “in the head” is part of the most powerful cognitive processes:

In the course of evolution, “the continued use and advancement of [language] would have reacted on the mind itself, by enabling and encouraging it to carry on long trains of thought. A complex train of thought can no more be carried on without the aid of words, whether spoken or silent, than a long calculation without the use of figures or algebra. It appears, also, that even an ordinary train of thought almost requires, or is greatly facilitated by some form of language”.

Why should the externalized forms of internal trains of thought be denied epistemic credit? (Rhetorical question.) Why should the use of emoji not be understood as an epistemic action? (Rhetorical question.)

Speaking of Darwin, he observed that “[t]he formation of different languages and of distinct species, and the proofs that both have been developed through a gradual process, are curiously parallel”. As we will see, these curiosities are paralleled in the development of emoji.

“We see variability in every tongue, and new words are continually cropping up; but as there is a limit to the powers of the memory, single words, like whole languages, gradually become extinct[…]:—‘A struggle for life is constantly going on amongst the words and grammatical forms in each language. The better, the shorter, the easier forms are constantly gaining the upper hand, and they owe their success to their own inherent virtue.’ To these more important causes of the survival of certain words, mere novelty and fashion may be added; for there is in the mind of man a strong love for slight changes in all things. The survival or preservation of certain favoured words in the struggle for existence is natural selection”

Darwin, The Descent of Man.

Indeed, consider the informal explanations for how it is that, as The New York Times reported, notwithstanding this time of plague, “Tears of joy prevailed as the most-used emoji in 2021, despite Gen Z’s stated contempt for it”:

“‘We did see a rise in the use of the virus emoji, but not in a way that even made it remotely into the most-commonly used emojis because we still had plenty to laugh about and plenty to cry about, whether it was because of the pandemic or not,’”. “‘Even in the midst of this massive global pandemic that preoccupied so much of our time,’ […]‘we still spent a lot of time wishing each other happy birthday or checking in or laughing about some new and unexpected element of this slow-burning weirdness’” (https://www.nytimes.com/2021/12/02/style/emojis-most-used.html).

Natural selection explains these phenomena rigorously: the “fittest” emoji that survive and thrive are those most expressive of human universals: the quiddities of our nature and the conditions it must confront. There will always be tragedy, comedy, and above all, irony. Darwin himself surveyed such universality in his Expression of the Emotions in Man and Animals, whose analyses of facial expressions could well illuminate interesting aspects of emoji. Consider the uncanny family resemblance between the illustration of “terror” that appears in Darwin’s book with its emoji analogue “fear”:


Let us assume then that the evolutionary “fitness” of an emoji is evidenced by (nota bene, not determined by) its frequency. Then it is natural to be interested in the frequency of emoji usage in the fashion of The New York Times:

“According to data from the Unicode Consortium, the organization that maintains the standards for digital text, nine of the 10 most-used emojis from 2019 (which was the last time they released data) also ranked among the top 10 this year. The red heart emoji held the No. 2 spot, and the tears of joy emoji ranked No. 1, despite members of Gen Z deeming it uncool (along with side parts and skinny jeans)[…]. According to data obtained from Twitter, tears of joy was the most tweeted emoji in 2020, but got bumped down to No. 2 this year, with the crying face taking its place. Tears of joy saw a 23 percent decline in usage from 2020 to 2021[…]. The syringe emoji jumped to 193rd place this year in terms of overall usage, compared to 282nd in 2019. The microbe also rose, from 1,086th in 2019 to 477th” (https://www.nytimes.com/2021/12/02/style/emojis-most-used.html).

We may predict that the usage of emoji conforms to “Zipf’s law”—a most curious and possibly spurious but consistently numerous power-law regularity in the frequency of word usage. Specifically, the frequency of a word tends to be approximately inversely proportional to its rank in frequency. Let f be the frequency of the word with the rank of r in a set of N words, then:

For instance, in a typical English corpus, the word with rank 1 is “the”, with a frequency of circa 70,000, and the word with rank 2 is “of”, with the frequency of circa 36,000: virtually exactly as Zipf’s law entails (i.e., 70000 × 1 ≈ 36000 × 2). The Zipfian characterization of word frequency can be visualized by plotting the log of word frequency against the log of word rank. By taking the log on both sides of the equation, supra, (log f = log C − log r), a perfect Zipfian fit would be a straight line with the slope -1. And indeed, Zipf’s law has been observed in vocabulary studies across languages and genres, and the log-log slope fit is consistently in the neighborhood of -1.0.

That being said, the status of Zipf’s law is controversial, which is manifest not only in word usage, but in myriad physical and social systems. Moreover, as Chomsky demonstrated, particular random letter generating processes can produce outputs that conform to Zipf’s law. For instance, if we redefine “words” as alphabets between any two occurrences of some letter, say, “e”, rather than spaces as in the case of written text, the resulting distribution may fit Zipf’s law even better. In the graph, infra, I show the Zipfian distribution of words (top) and pseudowords (bottom) in a typical English corpus. The lower line is plotted by taking “words” to be any sequence of letters between e’s (as Chomsky suggested). The two straight dotted lines are linear functions with the slope -1, which illustrate the goodness of the Zipfian fit.

Where would emoji fit? If Zipf’s law is characteristic of complex systems generally, including communication, as has been conjectured, then we may predict they would conform nomologically.

The conjecture that Zipf’s law obtains of any complex communication system enters into Oceanit’s work as a founding member of the Cambridge Institute for Exo-Language (CIEL). In the words of some of our collaborators:

“The concept of information panspermia introduced [as a solution] to the Fermi paradox, implies the transmission of […] life via compressed strings. The idea is based on the estimation […] that Earth organisms —including humans, and the entire terrestrial life up to bacteria— have common parts in their genomes. Hence the entire genome can be efficiently compressed[…]. After decoding, signals can represent themselves as traveling life streams. One can imagine that [life on some planets] is a result of such a transmitted package[…]. Conjecture: Once intelligent bit string signals will be compressed, we have to look at and reveal universal features of information compression which are absent in compressed signals originated via natural physical mechanisms[…]. It is natural to look for general features of communication systems, those which will presumably be present in meaning-conveying (non-random) messages, even if it is not (yet) known how to decipher this meaning” (Gurzadyan and Allahverdyan, “Non-random structures in universal compression and the Fermi paradox”).

Zipf’s law is conjectured to be one such universal, non-random feature of compressed signals. Oceanit and CIEL are working to test this information panspermia conjecture.

However, as I alluded, Zipfian regularities exist in prima facie non-linguistic systems. For instance, the great mathematician Benoit Mandelbrot conjectured a formal relation between Zipf’s (power-)law and fractal geometry. This theoretical conjecture has been corroborated empirically in Oceanitian work demonstrating the fractal structure of natural language syntax. Indeed, the fractal nature of languages—construed generally—is manifest in the language of molecular biology, as Oceanit has demonstrated in its work on RNA grammars (with applications for SARS-CoV-2 testing, etc.) and its LILI (life, information, language, intelligence) program. Zipf-like power-law patterns have been observed in the distribution of RNA (and protein) folds and beyond into oligonucleotide frequencies, the sizes of gene families (including pseudogenes), levels of gene expression, inter alia. As in the case of natural language, various and cogent explanations for these power-law patterns have been proffered, such as their mathematical relationship to scale-free networks, as could be predicted for metabolic pathways and protein interaction maps; in addition, there are models for how they could emerge in the evolution of protein families, all of which evince comparison to properties of words.

Consider the following graph, which plots the distributions of the number of occurrences of Pfam protein domains (blue squares) in the genome of the yeast Saccharomyces cerevisiae, and of words (red diamonds) in Shakespeare’s Romeo and Juliet, in both cases arranged in rank order from left to right. The most frequently occurring domains and words are labelled. In both cases (and in many other genomes and texts) the curves are good fits to Zipf’s power-law distribution relating the frequency to the inverse of the rank. It only remains to add emoji to the graph! (Could one tell the story of Juliet and her Romeo in emoji? I leave it as an exercise for the reader.)

Finally, something (or a “nothing”) that is overlooked in all the emoji analysis is sparsity: the fact that so much can be said with so little, and that everything that goes without saying or, deeper still, everything left unsaid, is perhaps the most profound aspect of the medium. It is redolent of Wittgenstein’s notion of the “mystical”: “There are, indeed, things that cannot be put into words. They make themselves manifest. They are what is mystical” (Wittgenstein, Tractatus Logico-Philosophicus). Perhaps emoji communicate something that cannot be put into words. Perhaps not.

(Perhaps not indeed, for I would reject the claim that some thoughts cannot be put into words; however, the argument is beyond the scope of this squib (but see Watumull and Chomsky, “A Minimalist Program for Artificial Intelligence” (In Press).)

Obviously, the typically unserious and even flippant emoji of today cannot be seen as mystical in the Wittgensteinian sense. However, as discussed in the course of this squib, it is in the nature of language and its external representations to evolve. We can imagine, therefore, that emoji could transcend their tawdriness as “Man” has transcended his “lowly origin”,  and arisen with “exalted powers”, “with all his noble qualities, with sympathy which feels for the most debased, with benevolence which extends not only to other men but to the humblest living creature, with his god-like intellect which has penetrated into the movements and constitution of the solar system” (Darwin, The Descent of Man).

Thus, in the future, we may see the mystical countenance of language literally manifest in the universal countenances of emoji, and limned almost perfectly in this Borgesian epitaph, which could well describe the sparse yet substantive—and arguably wabi-sabi—aesthetic of ideal emoji:

“Music, state of happiness, mythology, faces belabored by time, certain twilights and certain places try to tell us something, or have said something we should not have missed, or are about to say something; this imminence of a revelation which does not occur is, perhaps, the aesthetic phenomenon.”

Borges, “The Wall and the Books”

In summation, I submit that our use of emoji demonstrates that when used, emoji assimilate into our cognitive processes, functioning as external tools to process thought, analogous to the way we use paper and pencil to enable long-division. And, quite interestingly, the way we combine various emoji to convey completely different meanings is remarkably similar to Egyptian hieroglyphs. And in the same way that languages and hieroglyphs tend to evolve, consistent with Darwin’s theory of species evolution, so too do emoji—modulo the fact that the most frequently used emoji remained relatively unchanged, notwithstanding the influence of recent global events and cultural shifts. That emoji usage has remained substantially unchanged over the past two years, under pressure of a pandemic, is evidence that we have adopted emoji as a “complementary cognitive [specifically linguistic] artifact” (as opposed to mere decoration or novelty). Interestingly, and importantly, perhaps the most profound property of the truly linguistic nature of emoji is the fact that the implicit meanings with each character or combination thereof says vastly more than the characters themselves. We may reasonably conject that what remains unsaid conveys significantly more meaning than what is said.