The @ Symbol in Email: The Accident That Shaped How We Communicate

The @ symbol is so tied to email that we barely notice it. You type it every time you address a message or log into a service. But it wasn’t designed for that. Its path from obscure accounting notation to the centre of digital identity is a mix of accident, reuse, and one engineer’s need for a delimiter that wouldn’t appear in anyone’s name. Here’s how @ became the symbol of “where” in electronic mail.

Before Email: @ Had Other Jobs

The @ character has been around for centuries. Historians have found it in medieval manuscripts, where it was used as an abbreviation for “at” in commercial and notarial text—as in “12 barrels @ 3 ducats each.” It showed up on typewriter keyboards in the late 19th century and was used in accounting and commerce to mean “at a rate of” or “each.” So when Ray Tomlinson was looking for a character to separate a user’s name from the name of the host machine in the first networked email addresses, @ was already on the keyboard and already meant “at.” It was a natural fit: user at host.

Tomlinson was working on the ARPANET in 1971. He had built a program that could send messages between computers, and he needed a way to specify both the person and the machine. The separator had to be a single character that wouldn’t appear in usernames or hostnames. Most punctuation was ruled out: commas, periods, and slashes were already used in paths and addresses. Colons and semicolons had other roles in programming. The # and % symbols could appear in names or in system syntax. @ was rare in names and in system syntax, and it had a clear reading: “this user, at this host.” So he chose it. The first email address in the modern sense was something like username@hostname, and the convention stuck. He later said he didn’t remember the exact first address he sent—only that it was between two machines sitting side by side. The choice of @ was pragmatic, not symbolic; it just worked.

Why @ Won

Once ARPANET adopted the format, other mail systems did too. When the internet standardised on SMTP (Simple Mail Transfer Protocol) and the domain name system, user@domain was already the norm. So @ didn’t need to be “invented” for email; it was repurposed. Its prior meaning—”at”—made it readable. Its rarity in names and in other technical notation made it safe. And because it was a single character, it was easy to type and parse. No committee had to vote on it; one engineer made a practical choice and the rest of the world followed. RFC 5322, which defines internet message format, still specifies @ as the local-part separator. Billions of addresses today follow the same pattern Tomlinson used in 1971.

That’s a recurring pattern in tech: a small, local decision becomes permanent because it’s good enough and spreads. The @ in email is like the QWERTY layout or the width of railway gauges—once enough systems depend on it, changing it is costly. So we still type user@domain decades later, and the symbol has taken on a life beyond email. We say “at” when we read it aloud; we use it in handles and usernames on social platforms; we see it on signs and in ads. It has become shorthand for “address” or “contact” or “identity on the network.” No standards body had to mandate it; adoption did the work. Every new mail system, every new social platform that used “user@host” or “@handle” reinforced the same convention. The symbol earned its place by being there first and fitting the problem.

@ Beyond Email

As the web and social media grew, @ was borrowed again. Twitter (now X) used @ for mentions and handles: @username. That reinforced the idea that @ points to a person or an account—a destination. Other platforms adopted the same convention. So the symbol that once meant “at a rate of” in ledgers, then “user at host” in email, now often means “this account” or “tag this person.” The meaning shifted from location (at which machine) to identity (which account).

In many languages, @ has acquired a nickname. In Spanish it’s sometimes called “arroba,” an old unit of weight; in others it’s “at” or “monkey tail” or “strudel.” The variety of names shows how widely it’s used and how different cultures folded it into speech. But in code and in standards, it’s still the same character: one byte, one delimiter, one convention that started with a single design decision in 1971.

Why It Still Matters

Today @ does more than separate user from host. It’s the visual anchor for “this is an address” or “this is a handle.” When you see name@company.com or @username on a social feed, you don’t have to think about ARPANET or typewriters. You know what it means. That’s the mark of a successful standard: it becomes invisible. The @ symbol is one of the few pieces of 1970s internet design that every non-technical user recognises. It didn’t need a rebrand or a marketing campaign; it just needed to be useful once, in the right place, and then spread. The accident of Tomlinson’s choice gave the whole system a readable, typeable, parseable way to say “where” in electronic mail—and later, “who” in social and identity systems. Not bad for a character that started out meaning “at a rate of” in a ledger.

The Takeaway

The @ symbol in email wasn’t focus-grouped or standardised by committee. It was chosen because it was on the keyboard, it meant “at,” and it didn’t clash with existing syntax. That accident of history gave us a character that’s now inseparable from how we think about addresses and identity online. So the next time you type it, you’re using a piece of commercial shorthand that an engineer repurposed half a century ago—and that the whole internet still runs on.