Why I hate emojis

I have often voiced disdain for emojis, and my unhappiness that they are taking an ever-growing section of the Unicode space. Let me tell you why.

Emojis are pictographs. And pictographs suck.

Pictographs are symbols that derive their meaning from their appearance – they symbolize the thing they are a picture of. In contrast to pictographs we have phonetic alphabets and other abstract glyphs such as those used for math or punctuation. These abstract glyphs have a meaning that is entirely divorced from their shape.

Pictographs have one great strength, which is that they can be easy to learn. A picture of a tree means a tree. A picture of a bear means a bear. A picture of a mountain means a mountain. Since abstract glyphs have a meaning separate from their visual representation, they must be learned by practice and repetition. Ask any four or five year old whether they like pictographs or phonetic alphabets better and I’m sure they’ll generally go for pictographs.

That seems like a good advantage. Four year olds can understand simple pictographs while they still can’t understand even the simplest word written in a phonetic alphabet. However, that is the only advantage.

So here is a list of disadvantages:

• Pictographs are hard to write with. As a flimsy piece of evidence, cultures that use pictographs seem to have a stronger history of calligraphy. China, for instance→non-phonetic Chinese characters evolved from pictographs. It is difficult to draw things that are easily recognizable – just play a game of pictionary to see how hard it really is! Abstract glyphs, however, tend to be easy. The Latin characters are mostly quite distinguishable from each other, so little care needs to be taken to avoid making the letter ’b’ look like the letter ’g’. Since they are so easy to distinguish, even the chicken-scratch of a first or second grader is easily recognizable.

  Sure, you may say, they are hard to write. But nobody writes with pen and paper these days, everyone uses computers! Aha! My argument here is even easier. If you write by composing phonetic characters, it is easy to fit those characters on, say, a keyboard! I doubt that the typewriter ever would have been invented in a culture that relied on pictographs, but in a culture with a phonetic alphabet we enjoy the typewriter interface for our computers every day. Inputting a large number of pictographs on a computer, on the other hand, is rather unmanageable. The state of the art of emoji input is selecting (via finger press or mouse click) from a large selection on screen. While the average speed of typing is somewhere in the ballpark of 40 words per minute, it would be impossible to input pictographs nearly as fast, especially with a corpus of emojis that spans several pages.

  Writing with pictographs sucks.

• While pictographs for physical, visible things are easy, pictographs for invisible or abstract things are hard. What is the picture for love? A heart (♥), of course. But then how do we distinguish when a heart means a heart and when a heart means love? What about hate? Maybe an angry face? How do well tell the more nuanced difference between hate and anger? What should a pictograph for compound interest be? Probably a picture for the word compound with a picture for interest. What on earth should those pictures be?

  While grappling with the sheer number of abstract things to express and understand is difficult with both pictographs and phonetic writing, phonetic writing has the advantage of being connected to our spoken language, which we use from an earlier age, and which is more natural. So it is easier to amass a vocabulary for both the written and spoken word when these vocabularies are essentially the same. If there is no connection between your spoken word for something and your pictograph for it, you have to amass twice the vocabulary →I also have this problem with the profusion of obscure math symbols, but that is unrelated to pictographs..

  Also, there is additional complexity in the composition of pictographs. When you need to use two or more pictographs to represent something, what rules should guide that composition? For phonetic writing, the rules are the same as for speech.

  Additionally, there is a lot of cultural bias for what a picture should mean that makes pictographs even harder. Spoken and phonetically written words have a certain etymology, or history of where they came from. Pictographs can have the same thing, but it can be separate from the history of the spoken word. So you need much more cultural context to understand what people mean with pictographs than to understand what people mean with spoken or phonetically written words – even if the spoken/written words are slang.

  Trying to communicate anything worthwhile with pictographs sucks.

• Using pictographs, you need a massive library to be able to express worthwhile things. However, with composable glyphs such as phonetic characters, you only need a small hand full. The ASCII character set contains all the Latin letters and common English punctuation, and does so with fewer than 128 symbols. This means each ASCII character requires only 7 bits to store in a computer! Of course, we want to support other writing systems and glyphs as well, so Unicode is great. However, pictograph-based writing and emojis take a much larger part of the Unicode encoding space.

  Ok, so it takes up more of the encoding space. That’s not such a huge deal. But how about drawing those glyphs on your screen? The characters available in ASCII have a bagillion different fonts available for use, and it is not a terribly huge feat to create a new font for them. If you try to read a lot of emoji, however, you may find a lot of "no glyph found for this Unicode character" fallback glyphs. You know, they tend to be little boxes. So you tend not to have much choice in fonts with emoji, and there is far from universal support for them.

  What’s that, I complained about there being not enough fonts, and now I’m going to complain about there being too many? Yep. Since emoji glyphs are not universal, you have no guarantee that the picture you are sending is the same as the picture that will be received. And since pictographs derive their meaning from their visual form, that is kind of a big deal. The actual picture given by an emoji is subject to the ephemeral whims of fashion. Not only will your communication possibly be changed when it reaches an audience today, but it will almost certainly be changed when it reaches an audience only a few years from now.

  Trying to encode, decode and draw pictographs on computers sucks.

At this point, dear reader, you may say, "But emojis are only used to add small accents to otherwise entirely phonetic written English→Or some other language., so why does all of this matter?"

Well, perhaps it doesn’t. But I believe all of those problems exist even in the low doses of the occasional emoji. The original motivation for emoji→Or their predecessors, which I actually prefer for reasons of taste, emoticons. was to make up for the nuance of spoken language that is lost in writing. Are you writing something that you would have said in a light, joking manner? Add a smiley face! Are you writing something that needs to be taken seriously? Add a serious face! Oh, wait... actually, don’t. Are you writing something that shows that you are angry? Add an angry face! These don’t actually add any of the nuance of spoken voice, but a friendly smiley emoticon can occasionally go a long way in communicating friendliness in an internet where readers tend to assume hostility from any disagreement.

But aside from the handful of use cases like adding a smile or wink to coarsely communicate friendliness, joking, etc, what are they good for? What could you possibly communicate with a stock tiny cartoon of a snake in a message aside from simply the word "snake"? And if that is all you are communicating, why don’t you just write S-N-A-K-E? If you want to say anything interesting in a visual way about snakes, why don’t you just embed a picture of a snake in your message? If you are trying to add a more nuanced emotional backdrop or add any of the character that is lost by penning the spoken word, emojis will not help in a way that will be understood by readers. An emotional preamble explaining your intentions will be much easier to understand.

No intelligent and meaningful writing uses (or will ever use, I predict) emoji. So why go to all the effort to support vapid pictographic communication in our digital systems? Why not just improve support for markups that can include arbitrary embedded images instead? They say a picture is worth a thousand words. And often they are. But in the case of an emoji, I would rather have a few words any day →Ok, the real use of emojis is mostly girls being cutesy. Or advertisers in the subject lines of spam email. The real, real reason I hate emoji is because I resent them being in something technical that I and all technologists have to work with (the Unicode spec, the vagaries of printing/displaying text, etc). I think it lowers the intellectuality of the communications of teenagers who use them rather than using words. Emoji only supports poor communication and the arms race to make spam email more annoying..