In 2011, Apple became the first big tech company in the West to visibly embrace emoji. The detailed, glossy symbols that appeared that year on the iPhone’s on-screen keyboard were a far cry from Shigetaka Kurita’s lo-fi efforts and they went on to become the de facto standard for modern emoji design. But though Apple holds the emoji 👑, it was Gmail, Google’s email service, that had first dragged emoji out of Japan and onto the world stage. And drag it had to, for emoji did not come quietly.
In the mid-2000s, as Google sought to expand its reach in Asia, it prepared to make Gmail, its email service, available to users in Japan. Emoji were as unfamiliar outside their native country as they were beloved inside it,* but Takeshi Kishimoto, product manager for Google in Japan, knew that a successful launch would depend on their inclusion. His bosses agreed in principle but balked at one symbol in particular: Takeshi was adamant that Gmail must include a poo emoji.2 Promptly, the 💩 hit the fan.
The poop emoji was based on a character from a 1980s anime series called Dr. Slump. “Poop Boy” was one of a parade of poo-related characters in the series, appearing alongside “Manure Boy”, “Bird Poop Boy”, “Old Man Poop” and “Soft-Serve Ice Cream Boy” (whom the others mistake for a pile of faeces), and he went on to appear both in other anime series and a set of related video games.3 Poop Boy’s digital alter ego was introduced in 2000 by KDDI AU, one of NTT DOCOMO’s competitor networks,4 and it was embraced by younger users as a synonym for the word unchi, a childish exclamation meaning “poop” or “shit”. It was the visual equivalent of the word “doo-doo”, in other words, with all the earnest utility and gleeful mischief that implies.2 Users loved it — and Takeshi could prove that they loved it, showing his bosses a study that ranked ‘💩’ as one of Japan’s most popular emoji.5 His appeal to the data won the day and the poop’s place in Gmail was assured.
But Google’s engineers faced another problem when implementing ‘💩’ and its sibling emoji. Each of Japan’s mobile networks supported more or less the same set of emoji as the others, but the binary numbers that identified those emoji differed from one company to the next. Send a ‘📺’ from your NTT cellphone to a friend on KDDI AU and they would see ‘💡’; your colleague on the SoftBank network, ‘🌄’. The result was what the Japanese called mojibake, or garbled characters.6 It was left to Google to cajole emoji into working across all of Japan’s mobile networks and Google, in turn, sought the help of an organisation called the Unicode Consortium.
Unicode’s roots stretch back to the 1980s. At that time, most computers understood only a limited set of characters, and the characters that they did understand were often specific to a single country or language. Opening a file that had been created on a computer in a different region often resulted in a mess of misinterpreted text — mojibake before emoji ever existed. Finally, in 1987, the leaders of multilingual computing projects at Apple and Xerox joined forces to develop a single universal character set they hoped would replace the era’s cacophony of regional standards. They called it “Unicode”.7
Since its incorporation in 1991, the Unicode Consortium’s cabal of engineers, linguists and government apparatchiks have worked to codify more or less all forms of writing known to humanity.7 The group’s achievements include such noble works as making the ancient Linear B alphabet available to scholars across the world; preserving archaic Egyptian Coptic for posterity; and enabling once-marginalised scripts such as Cherokee to flourish online. All this has been done by asking and answering the same question, over and over again: what number should be assigned to this character, that character, and the one after that?8
Now, fifteen years after the consortium’s founding, Google knocked on Unicode’s doors and asked these éminences grises of computerised language: “What number should we use for ‘💩’?”
Unicode was unfazed. In fact, the organisation had unwittingly made a head start on the task of encoding emoji thanks to a hangover from the old days of metal typesetting.
In part two we saw how some of the more flamboyant characters to appear on the printed page were under pressure from both cultural and technical factors. Printers and book buyers preferred a neat, clean page with few adornments and, upon its arrival in the late nineteenth century, the typewriter’s utilitarian keyboard made it incrementally harder to use such characters in the first place.
But that is not the whole story. Even under the combined assault of the QWERTY keyboard and the minimalist aesthetic of the modern book, special typographic characters never really went away. Many such marks survived in the typographic priest holes called “pi fonts” — collections of symbols too niche for mainstream typefaces but too useful to die out entirely. The origins of the term “pi font” are elusive, but it may come from the phenomenon of “printer’s pie”, where an inattentive worker knocks a page of movable type onto the floor in a jumbled mess.9 If true, it would be an appropriate label: the typical pi font is home to an oddball mix of mathematical and technical symbols (∑, √, ƒ, %, ‰, º and more), currencies ($, ¥, £ and so on), fractions (½, ⅓, ¼), and a generous smattering of ancient and medieval characters (¶, ☞, ❦).10
Some pi fonts harboured even rarer characters, as can be seen in the 1935 edition of Madeline Brandeis’s book Carmen of the Golden Coast.11 These are honest-to-God typographic smileys, printed thirty years before Harvey Ball got the credit for inventing them and sixty before Shigetaka Kurita immortalised them in pixels. Why are they here? As with typefaces in general, individual pi characters were occasionally commissioned for use in a particular work, and this may be how Carmen’s singular smileys came about. Certainly, this is the only pre-emoji printed work I’ve come across that contains ‘🙂’ and ‘😭︎’ as individual glyphs, which suggests that these were one-offs. Even Little Tom of England, a similar book by the same author, published a year earlier, lacks them.12 But tempting as it is to draw a line of descent from Brandeis’s printed book to Kurita’s electronic icons, the evidence is not there to support it.
Even so, pi fonts can rightly claim a role in emoji’s emergence on the world stage. When Apple unveiled its LaserWriter printer in 1985, as a companion to 1984’s Macintosh computer,13 it came with a set of built-in fonts selected by none other than the company’s chairman, Steve Jobs. Three of those fonts came from the International Typeface Corporation, or ITC — a big win for the then-young type foundry — and two of those ITC fonts were products of a renowned type designer named Hermann Zapf. The first, Zapf Chancery, was designed to convey a handwritten feel (it was the only such “script” font that Jobs could bear to put on the Mac); the second, Zapf Dingbats, was a good old-fashioned pi font containing hundreds of miscellaneous symbols such as stars (✯), arrows (➽), crosses (✞), snowflakes (❄) and aeroplanes (✈).14 Together, the Mac and the LaserWriter brought high-quality printing to the masses and Zapf’s typefaces became mainstays of desktop publishing.
In 1991, then, when the Unicode Consortium published their first stab at a standard character set, a keen-eyed reader† would have noticed the inclusion of a sequence of characters taken directly from Zapf Dingbats. To create Unicode 1.0, the consortium had simply selected what it considered to be the most common extant character sets and typefaces, resolved the most obvious conflicts, and unified them into a single all-encompassing character set.15 Zapf Dingbats’ hallowed place in the Apple ecosystem and beyond meant that it was one of those industry standards, and so in it went. Another section in Unicode 1.0, “Miscellaneous Dingbats”, augmented Zapf’s symbols with yet more glyphs such as chess pieces (♛), warning symbols (☣, ☢, ☠), musical notes (♫, ♬), playing card suits (♠, ♣, ♥, ♦) — and a pair of smiling and frowning faces (☺, ☹).16
Thus, when Google came to Unicode a decade later, looking to standardise a disparate collection of emoji, Unicode had seen it all before. Aeroplanes? Hearts? Pointing hands? Smiling faces? Why yes, said Unicode, we’re way ahead of you.
Emoji’s admission into the Unicode standard, progressing as it did through layer after layer of discussion, procedure and committee, was slow but steady. The first formal step was taken in 2006, when Unicode co-founder Mark Davis (late of Apple; now at Google) proposed that the organisation take a look at Japan’s popular picture-writing symbols, explaining that:
There are a number of symbol sets that are in widespread use, but currently can only be mapped to private use characters on input. The UTC [Unicode Technical Committee] should consider whether or not it would be useful to encode these, or some subset.17
Davis did not mention emoji explicitly, but his meaning was clear and the ball was set rolling. A first draft of a unified emoji character set followed in 2007,18 and in 2010, the publication of Unicode 6.0 made it official: 722 emoji were added to the more than 100,000 other characters then in the standard.‡ Thanks to Zapf Dingbats and the other pi fonts on which Unicode had been founded, 114 of those emoji were already present.19
Finally, Google could add emoji to Gmail without worrying that its users would suffer from the dreaded mojibake. And where emoji went, so followed the poop: Gmail’s unchi emoji was an evocative, faceless turd surrounded by circling flies, its scatalogical impact blunted only by its diminutive 15-pixel-square dimensions.2 Moreover, the poop emoji and its kind were now free to roam the internet at large. First Apple, then Microsoft, Facebook, Twitter and many others followed Google in adding support for emoji to their applications, web pages and operating systems. Emoji had gone global, and it has not looked back since.
- “Search for ‘emoji’ Between 1999 and 2006”, Google Scholar. ↢
- Lauren Schwartzberg, “The Oral History Of The Poop Emoji (Or, How Google Brought Poop To America)”, Fast Company, 2014. ↢
- “Poop-Boy”, Dragon Ball Wiki. ↢
- Caitlin Harrington, “Origin of a Feces: A Not-So-Brief History of the Poop Emoji”, Wired, 2017. ↢
- Jeff Blagdon, “How Emoji Conquered the World”, The Verge, 2013. ↢
- Ritchie King, “Will Unicode Soon Be the Universal Code? [The Data]”, July–2012. ↢
- Laura Wideburg, “Early Years of Unicode”, Unicode.Org, 1998. ↢
- “What Is Unicode?”, Unicode.Org. ↢
- Richard Eckersley, “Pi Font”, Glossary of Typesetting Terms, 1994. ↢
- “Pi Characters”, PrintWiki. ↢
- Madeline Brandeis, Carmen of the Golden Coast, 1936. ↢
- Madeline Brandeis, Little Tom of England, 1935. ↢
- Jason O’Grady D, “Technology Timeline”, in Apple Inc., 2009, 72-75. ↢
- Ilene Strizver, “The Story Behind Zapf Chancery”, Fonts.Com. ↢
- The Unicode Consortium, “1.0 Introduction”, in Unicode Standard, Version 1.0, 1991, 1-6. ↢
- The Unicode Consortium, “3.2 Symbols”, in Unicode Standard, Version 1.0, 1991, 72-97. ↢
- Mark Davis, “L2/06-369: Symbols”, Unicode.Org, 2006. ↢
- Kat Momoi, Mark Davis, and Markus Scherer, “L2/07-257: Working Draft Proposal for Encoding Emoji Symbols”, Unicode.Org, 2007. ↢
- Mark Davis and Peter Edberg, “Unicode Technical Standard #51: Unicode Emoji (version 11.0)”, Unicode.Org, 2018. ↢
- A search for scholarly papers from the years 1999–2006 containing the word “emoji” returns only a handful of articles in English.1 ↢
- And a persistent one: the standard spanned 682 printed pages and contained more than 28,000 characters.15 ↢
- More followed in 2012 and 2014, and the yearly cadence of new Unicode releases has since morphed into “emoji season” — but more on that in later parts. ↢