Emoji Don't Mean What They Used To

The pictorial language has moved away from ideography and toward illustration. It’s lost some expressive power in the process.

Iraqi Kurds use emoji masks during a 2017 protest. (Getty)

As of last week, there are now 3,053 emoji, counting the 230 just approved for this year’s cohort—yes, the icons now get annual releases, like Microsoft Word or tax returns.

This is too many emoji. We now have icons representing people with disabilities—an admirable step toward digital representation— but we also have badgers, myriad types of rail, fingers folded in every style, superheroes, and genies. This year, we got garlic and a yo-yo.  Every new addition makes finding the right icon harder.

And still, not every emoji represents every possible identity. There are no redheaded brides, for example. There are no white men with brown hair and beards. The new emoji include hand-holding couples of different skin tones, allowing interracial couples their own mobile pictograms at last. But adding different skin-tone options for each family member in multi-person groups, such as the family (👪), would result in 4,225 permutations. This year’s class of emoji adds stand-alone gender-neutral options, but those aren’t available as professions like doctor or astronaut. Unless emoji get coupled to Bitmoji-style avatar creators, they will always leave some people out. And even then, who wants to create a new avatar for every individual they want to depict in a quick message?

As a white man whose identity is often the default in emoji, let me say explicitly that increasing diversity in the globe’s favorite pictorial language is a good thing. But all together, emoji are becoming more specific and less flexible as more icons appear. That shift doesn’t just add more choice among emoji; it also changes their semiotic function. Over time, the visual language has shifted away from abstract, ideographic uses and toward specific, illustrative ones.

A selection of the original emoji (NTT Docomo)

The original emoji were created for Japanese cellphones by the telecom NTT Docomo in 1999. Those emoji worked as pictograms or ideograms. Pictograms, such as the train or cigarette in the original set, work like international signage—they convey meaning by resembling an object. Ideograms are symbols that represent ideas or concepts rather than objects themselves—a circle with a line through it (🚫) to indicate prohibition, for example. Many emoji are hybrids of ideograms and pictograms. The heart or the snowman, for example, aren’t typically used to depict cardiac organs or winter yard sculptures. Instead, they signal love or coldness, respectively. The 1999 emoji were small and low-resolution, too: 12 pixels square, in a single color. That helped them work like airport signage rather than like avatars.

Pictograms (including ideogrammatic ones) are powerful because they are specific but flexible. The train can represent a light-rail line, a subway, a toy, and so on. A snowman can mean a literal snowman, or a warning that it’s cold out, or even a gripe about the office thermostat. The pleasure, and power, of emoji arises from the ambiguity inherent in picto-ideographic writing.

That power continues with today’s higher-resolution versions. A skull (💀) almost never means that the speaker has a braincase in hand, Hamlet-like, but rather offers an ashen reaction or a lol, I’m dead sentiment. An emoji originally designed to signify an Eastern bow of greeting or politesse (🙇‍♂️) takes on the more abstract meaning of mild subjugation or psychic deflation in the West. Fire (🔥) could mean a campfire or house fire, but more often it suggests enthusiasm, ferocity, or even spice. Eggplant (🍆) could denote a nightshade, but more likely it suggests, well, something else. These and other meanings are possible because the emoji function primarily as ideograms.

But as emoji have become more specific in both their appearance and their meaning, their ideographic flexibility has eroded. Consider two versions of the cocktail emoji pictured below:

NTT Docomo’s 1999 cocktail emoji (left) and Apple’s current rendition (right)

The 1999 version mostly bears ideographic meaning. It suggests “cocktail” in the abstract, and you can imagine using it to suggest that it’s time for drinks, or to indicate that you’re waiting at the bar, or to say that you’ve had a few drinks and shouldn’t drive, depending on the context. It works like a sign, not artwork.

The contemporary one can still function that way, at times. But it’s so detailed and specific that its individual utility wanes. It’s not an icon for a cocktail, but a picture of a martini (a dirty one, with an olive, no less). That’s been the trend in emojiland: The cocktail has been joined by a beer (🍺), red wine (🍷), whiskey (🥃), and even a mai tai (🍹), for example. There are clinking glasses of both beer (🍻) and champagne (🥂). More granularity and specificity offer more choice, but those choices are no longer ideographic; they are pictographic at best, and perhaps merely illustrative. “I propose a tumbler of whiskey,” or “I salute you in the manner that a formal-event toast implies.” Counterintuitively, all these emoji are less applicable because they contain more information.

The drive to offer more detailed information is behind most appeals for additions to the emoji character set, so that condition is likely to amplify. For example, this year, a blood drop will be added to the emoji set. The Unicode Consortium considers flexibility of use as a part of its review of applications, and the blood drop has the potential to signify all manner of hematological activities, from donating plasma to nosebleeds.

But the design rankled some, who had hoped for an emoji explicitly designed to depict menstruation. The humanitarian nonprofit Plan International UK had held a contest to that end in 2017, and submitted the winner, a pair of “period pants” (underwear emblazoned with a blood drop) to the Unicode Consortium, which passed on the offer (a follow-up, the blood drop, was approved). In a riposte of the more generic design, Slate’s Shannon Palus lamented that the blood drop “will not be as useful for … symbolizing the thing it was designed to represent and normalize: periods.”

It makes sense that emoji should strive to cover the gamut of human experience; more than half of human beings menstruate at some point in time, so that’s a good place to exert effort. But more specificity means less flexibility. That is, an emoji that shares other possible meanings, among them menstruation, is assumed by Palus and others to be a less desirable design choice than one with a singular, fixed meaning. That idea might or might not have political merit, but it does represent a shift in the way emoji have been conceived, approved, and used since the iPhone made them globally popular in 2010. The assumption that more numerous, more specific emoji are automatically better seems to be spreading, too. Appeals in the form of there’s no emoji for … have almost reached meme status.

These factors have changed the way emoji get created, selected, and used. Writing about the rise of the dumpling emoji in Fast Company, Harry McCracken explained how the consortium debates the global fallout of new character designs, including the implications of specific forms of dumplings (gyoza? ravioli? pierogi?). The result (🥟), designed by Yiying Lu, was a “a romantic ideal,” according to McCracken, “aiming for something that could cross cultural borders.”

The dumpling amounts to a success for pictography over mere illustration (even if some platforms don’t preserve its generic appearance). But it probably offers limited applicability compared with more ideographic emoji. The concept of global dumplingdom doesn’t afford the expressive flexibility of a toothy smile (😬), a siren (🚨), or a clamp (🗜).

A preference for specifics drives the appeals for diversity in emoji, too. Plan International UK staked a claim in the blood-drop emoji once it was announced: “We are actually getting a #PeriodEmoji,” for example. And before this year, only a generic, international icon of a wheelchair (♿️) represented disability in the picture set. That does seem insufficient. But that insufficiency isn’t necessarily caused by the quantity of disability emoji, or even by their overall specificity. It comes from a shortage of ideogrammatic power. To have only the equivalent of an airport sign for the subject feels like a slight on the varied and unique experiences of disabled people. And yet, the Unicode Consortium has responded to that shortfall by adding pictographic specificity: two types of wheelchairs, a service dog and a guide dog, a mechanical arm and leg, and so on.

Don’t get me wrong: All those new disability-friendly emoji are great! But they offer more evidence that emoji are transforming into a large catalog of fixed portraits, rather than a smaller set of flexible ideograms. That shift doesn’t just add to emoji; it also changes how they work.

The awkwardness of the interfaces used to access emoji amplify that change. Overwhelmed by choice, we’ve become more tempted to type in a word and have the device offer matches, as some emoji interfaces allow. That’s also how some text-entry systems for nonalphabetic languages work. But unlike logograms—pictures that represent a word or phrase, like those used for Chinese characters—ideographic emoji thrive when their meanings remain ambiguous. Matching icons to words encourages fixity of meaning, especially as it becomes harder to find any single emoji by scrolling.

Emoji are transitioning from pictograms to pictures. That change offers some obvious benefits, like the ability to create images that better represent a broader set of individuals and their experiences. It also shifts emoji’s function toward specificity and away from abstraction. Emoji is humankind’s weirdest and most successful ideographic language. If it is to become an illustrative one instead, that’s a revision worth discussing with words, not just celebrating (🎉), or lamenting (🙅), with pictures.