Three words, used interchangeably by almost everyone, describing three genuinely different things. Here’s the short version:
| Example | What it is | Origin | |
|---|---|---|---|
| Emoticon | 🙂 | A face built from ordinary keyboard punctuation, read sideways | USA, 1982 |
| Emoji | 😀 | A single picture character with its own Unicode number | Japan, 1999 |
| Kaomoji | (^_^) | A face built from many characters, read upright | Japan, 1980s |
The differences aren’t just historical trivia. They decide which ones survive in a game name field, which ones look different on your friend’s phone, and which ones a screen reader can handle.
Emoticon: the one you build yourself
An emoticon is a face assembled from punctuation you already have. A colon, a hyphen, a bracket. Tilt your head left and 🙂 becomes a smile.
It started on a university message board. On 19 September 1982, at 11:44 in the morning, a computer science professor at Carnegie Mellon named Scott Fahlman suggested using 🙂 to mark posts that were jokes, and 🙁 for posts that weren’t. Guinness treats that post as the birth of the digital emoticon. The word “emoticon” itself was coined later by someone else entirely.
The message was assumed lost for two decades, then recovered from old backup tapes in 2002.
What matters technically: an emoticon is just ordinary text. 🙂 is three characters — a colon, a hyphen, a close bracket — and nothing about it is special. It renders identically on every device ever made, works in any input field, and can’t break.
That’s why emoticons still show up in places nothing else survives: :) in a game name, :D in a form field that rejects everything else.
Emoji: the picture characters
An emoji is a single character that your device draws as a picture. 😀 isn’t a face made of parts. It’s one character, U+1F600, with its own number in the Unicode standard, exactly like the letter A has one.
The word has nothing to do with “emotion.” That resemblance is a coincidence that fools nearly everyone. Emoji is Japanese: 絵 (e, picture) plus 文字 (moji, character). Picture-character. The English word “emotion” comes from Latin and isn’t related at all.
The famous set came from NTT DoCoMo in 1999. A designer named Shigetaka Kurita was working on i-mode, an early mobile internet service whose emails were capped at 250 characters. With that little room, he needed a way to carry tone compactly, so he drew 176 tiny images on a 12×12 pixel grid, borrowing from manga, Zapf Dingbats and the emoticons already in use. That original set now sits in the collection of the Museum of Modern Art in New York.
Worth noting for accuracy: Kurita’s wasn’t strictly the first. J-Phone put 90 emoji on a handset in 1997, and Kurita has acknowledged as much. His was simply the set that spread, because i-mode shipped across many devices while the earlier phone sold poorly.
Emoji entered Unicode properly in 2010, which is what made them work across brands. As of Emoji 17.0, released on 9 September 2025, there are 3,953 emoji in the standard — a figure that counts skin-tone and gender variants, flags and the components used to build them.
Kaomoji: the Japanese faces you don’t tilt
Kaomoji are the elaborate ones: (^_^) ಠ_ಠ ¯\_(ツ)_/¯
The name says what they are — 顔 (kao, face) plus 文字 (moji, character). Face-characters. They emerged on Japanese computer networks in the 1980s, in parallel with Western emoticons, and they solve the same problem in a completely different way.
The key difference is orientation. A Western emoticon is drawn sideways and needs you to tilt your head. A kaomoji is drawn upright and doesn’t. The eyes are where eyes should be.
There’s a reasonable theory about why: Japanese typing already had a large character set within easy reach, so there was no need to economise on characters the way an ASCII keyboard forces you to. With ^ ω ಠ ͡ and hundreds of others available, you can draw a face properly instead of suggesting one with three keystrokes.
That’s also why kaomoji are long. ¯\_(ツ)_/¯ is nine characters. (╯°□°)╯︵ ┻━┻ is twelve — and the bits that make up the flipped table are box-drawing characters, the same ones once used to draw borders in text-mode software. The ツ in the shrug is simply the Japanese katakana letter tsu, borrowed because it looks like a face.
The difference that actually matters
Understanding what each one is made of predicts how it behaves:
Emoticon :-) | Emoji 😀 | Kaomoji (^_^) | |
|---|---|---|---|
| Built from | Keyboard punctuation | One dedicated character | Many characters, often non-Latin |
| Looks the same everywhere? | Yes, always | No — each platform draws its own | Usually, but rare characters can fail |
| Survives in game name fields? | Yes | Usually stripped | Often, if the field allows the characters |
| Costs against a character limit | 3 | 2 or more | 5–15 |
| Screen reader handling | Read as punctuation | Read by name, e.g. “grinning face” | Read character by character — messy |
Emoji look different on every platform
Apple, Google, Samsung and WhatsApp each draw their own artwork for the same character. You send one thing; your friend sees another interpretation of it. In the early days this was far worse — before Unicode standardised them, a smiling face sent between networks could arrive as something else entirely.
Emoji cost more than they look
😀 is one character, but most software stores it as two units and counts it that way. Add a skin tone and 👍🏽 becomes two characters and four units. A family emoji 👨👩👧👦 is actually seven characters joined by invisible connectors — eleven units against your limit. This is why a bio with a few emoji hits Instagram’s 150-character cap sooner than it should. (Here’s the underlying reason characters and units differ.)
Which one to use where
Game name fields — emoticons and kaomoji. Most games strip colour emoji from names outright, but accept punctuation and katakana. A trailing ツ survives almost everywhere.
Instagram and TikTok bios — any of the three. Emoji are safest and most legible at small sizes; kaomoji give more personality but eat characters.
Usernames and @handles — none of them. Those fields accept only letters, numbers, periods and underscores.
Professional messaging — emoticons or nothing. :) reads as tone; 😂 reads as informality, which may or may not be what you want in a work email.
Anywhere accessibility matters — emoji, used sparingly. Screen readers announce them by name, which works. A kaomoji gets read out as a string of individual character names, which is unpleasant to listen to.
Grab them
Browse the full sets: emoji, emoticons and kaomoji, all copy-and-paste.
Common questions
Is 😀 an emoji or an emoticon?
An emoji. If it’s a picture drawn by your device, it’s an emoji. If it’s punctuation arranged to look like a face, it’s an emoticon.
Is ¯\_(ツ)_/¯ a kaomoji?
Yes. It’s read upright, it’s built from multiple characters, and it follows the Japanese style.
Why do my emoji look different on Android and iPhone?
Each company draws its own artwork for the same character. The character is identical; the picture isn’t.
Why does an emoji show as a box or a question mark?
The device has no artwork for that character — usually because the emoji is newer than the phone’s software. It arrived fine; the phone can’t draw it.
Which is oldest?
The Western emoticon, 1982. Kaomoji followed in Japan during the 1980s, and emoji arrived on phones in the late 1990s.
Are emoji the same as smileys?
“Smiley” usually means the yellow round face specifically. Emoji covers everything — food, flags, animals, objects.