History

This Day in History — The Digital Smiley Is Born :-)

On this day in 1982

On September 19, 1982, at 11:44 AM, Scott Fahlman posted :-) and :-( to a CMU bulletin board. Guinness calls it the birth of the digital emoticon. He never made a dime from it.

The Off-Key Bard stares at an early computer terminal where people have discovered one of the oldest problems in online communication: without a face or tone of voice, sometimes nobody can tell when you're joking.

"The solution didn't require faster computers or better graphics. It required three characters: :-)"

On this day in 1982, at exactly 11:44 in the morning, computer scientist Scott E. Fahlman posted a short message to the Carnegie Mellon University Computer Science Department's bulletin board — a bboard, in the parlance of the era — and proposed two sequences of characters that would spread from one university to the next, eventually circling the globe.

The message read, in its entirety:

"I propose that the following character sequence for joke markers: :-) Read it sideways. Actually, it is probably more economical to mark things that are NOT jokes, given current trends. For this, use :-("

It was, Fahlman later acknowledged, a casual suggestion dashed off in a hurry — missing a few words, as he noted. He had no idea he was starting anything.

The Thread That Started It

The conversation that produced the proposal had begun three days earlier, on September 16, when CMU graduate student Neil Swartz posted a physics brainteaser to the bulletin board: imagine a candle in a falling elevator — what happens to the flame, and what happens to a drop of mercury on the floor? A few hours later, Howard Gayle responded not with an answer to the puzzle but with a deadpan "WARNING!" post, written in mock-urgent tones, as if the elevator scenario were a genuine safety hazard.

Several people took Gayle's warning seriously. A confused thread ensued. The bulletin board community began discussing the problem of sarcasm and irony in written communication — how, without facial expression or tone of voice, an obvious joke could be read as a genuine statement. Various proposals were offered for how to mark joke posts. Some involved elaborate notation; some were never seriously meant. On September 19, Fahlman offered his.

The two symbols were clever in their simplicity. Both :-) and :-( were composed entirely of characters available on any standard keyboard. Both were recognizable as faces — but only if you tilted your head sideways and squinted a little. That slight perceptual shift was, Fahlman reasoned, effort enough to signal that something in the post was not entirely straight. The instruction "Read it sideways" became necessary.

How It Spread

Within Carnegie Mellon, the symbols caught on quickly. By November 1982 — just weeks later — CMU professor James Morris had compiled an expanded list of emoticons and forwarded them to a colleague at Xerox's Palo Alto Research Center, PARC. That transmission from one research network to another was, effectively, the beginning of the emoticon's spread into the broader computer science community. From research labs it went to early internet forums and from there to email, instant messaging, and eventually the entire internet.

The original message itself seemed to vanish in the process. By the early 2000s, when Fahlman told the story at conferences and in interviews, no one could locate the actual post. Then, around the twentieth anniversary, a group of CMU researchers undertook what Fahlman called the Digital Coelacanth Project — named after the famously "extinct" fish that turned up alive off Madagascar in 1938. They located old backup magnetic tapes from an October 1982 DEC VAX-11/780, found hardware still capable of reading them, decoded the obsolete formats, and on September 10, 2002, Jeff Baird successfully recovered the original message. The date, the time — 11:44 AM — and the exact wording were confirmed.

The Guinness Book of World Records now recognizes Fahlman's September 19, 1982 post as the birth of the digital emoticon.

Fahlman, reflecting on what he started in his 40th anniversary essay in 2022, noted that the word "emoticon" — a portmanteau of "emotion" and "icon" — was coined later by someone he never identified. He did not coin the term. He also noted, with characteristic dry humor, that he had never made a dime from the invention, and that he expected the first line of his obituary to mention it.

Emoticons, Emoji, and What Came Next

The emoticon Fahlman proposed was a text-based construction — ordinary ASCII characters, typed in sequence, forming a recognizable image only when tilted sideways. Emoticons spawned a vast extended vocabulary: ;-) for winking, :-P for sticking out a tongue, :-D for a wide grin, :-| for blankness. Different communities developed different variations.

The emoji that now fill our phones are something technically distinct. Shigetaka Kurita, working for Japanese mobile carrier NTT DoCoMo, helped develop the first influential emoji set in 1999 — small 12-by-12-pixel pictographic grid images designed for mobile pagers, each character encoding an emotion or concept as a single glyph rather than a sequence of ASCII characters. When Unicode standardized emoji across platforms and Apple added an emoji keyboard in iOS 5 in 2011, the pictographic vocabulary exploded into the hundreds of thousands.

Both addressed the same underlying problem: human communication carries far more meaning than text alone can convey. We signal irony with a raised eyebrow, warmth with a smile, sarcasm with a flat tone. Writing strips all of that away. Fahlman's three characters were a minimal workaround — a small, lateral-thinking solution to a problem that no amount of faster processors would have solved on its own.

"A colon, dash, a bracket bent,
Could show just what the writer meant.
Three symbols turned and read sideways —
And gave the digital world a face."

History reminds us: the greatest breakthroughs in computing aren't always about processing power. Sometimes they are about recognizing that the machine is transmitting text with perfect fidelity — while the human meaning gets lost in translation.

Sources