Emoji interoperability (or its lack)

Unicode characters ought to have a specific denotation, even if their exact appearance depends on the font. A letter, a punctuation mark, or a Chinese ideograph should have the same meaning to everyone who reads it. There are problems, of course. There’s no systematic difference in appearance between A, the first letter of the Roman alphabet, and Α, Alpha, the first letter of the Greek alphabet. (However, when I had my computer read this article aloud to me for proofreading, it pronounced the latter as “Greek capital letter alpha”! Nice! It also pronounced the names of the emoji in this article, except the new ones in Unicode 11.0.) In some fonts, you can’t even tell the lower case letter l from the number 1 without looking carefully. This problem allows homograph attacks and “typosquatting.”

But the worst problem is with the Unicode Consortium’s great headache, emoji. These picture characters have just brief verbal descriptions in the Unicode standard, and font designers for different companies produce renderings that have vastly different connotations. Motherboard offers a sampling of the varied renderings. Here’s the “grimacing face” from Apple, Google, Samsung, and LG respectively.

Four renderings of grimacing face emoji

Emojipedia has even more variants and asserts that on Snapchat “this emoji next to a friend denotes that you send the most snaps to the same person that they do.” All that from a cartoon picture of a face!

Here’s what it looks like on the device you’re using right now: 😬

Even different applications on the same device could render emoji differently, depending on the font they use. A font could have an incomplete emoji repertoire, especially if it hasn’t caught up with the latest Unicode release. Unicode 11.0 lists 161 new emoji. You can now present a supervillain 🦹 who sinks his teeth 🦷 into a bagel 🥯, but that may or may not show up as anything but a placeholder on your browser, depending on how recent your font is.

Sometimes cultural differences make emoji mean different things to different people, even without any major variation in appearance. The “sleepy face” emoji 😪 looks like a sad, crying face to most Westerners, but apparently people who watch enough anime get it. The Washington Post mentions this and other emoji, citing polls that show how differently people interpret them.

The Unicode Consortium has done some work on glyph recommendations to improve the situation, but it may have given the task up as hopeless. Most of the references I can find are dated 2015.

Before 2010, the situation was worse. Each manufacturer used the Unicode private space to define its own emoji encodings, independent of anyone else’s. But it seems like a hopeless job. The only lesson to get out of it is: Use emoji with caution. Try not to use them at all for serious business communications. They’re for fun, not for serious communication, at least around here. The Japanese, as I understand it, take emoji more seriously. Maybe they can make sense of it all.

Comments are closed.