The Surprising Origins of Punctuation

You probably went your entire reading life without wondering where the period came from. The marks on a page are so ordinary they vanish, the way furniture in a familiar room vanishes. But every one of them is an invention. Each one was argued about, resisted, standardized, and forgotten. Together they form the longest-running collaborative typography project in history, and most of it was finished before anyone wrote the rules down.

The story is older than printing, more chaotic than school grammar suggests, and quietly fascinating once you start to notice that nothing on the page is inevitable.

Before punctuation

The earliest Greek and Latin manuscripts had no spaces between words. Scriptio continua, it was called. A line of text was a single ribbon of letters, and reading it required parsing aloud. Silent reading was unusual until well into the medieval period; reading was a performance, and the reader was expected to find the rhythm by sounding out the words.

This was not, as we tend to assume, an oversight. Spaces between words required ink, parchment was expensive, and a trained scholar reading aloud could parse the stream as fast as he could speak. The Romans had marks for emphasis and quotation, but no commas, no periods, no question marks. The earliest Greeks did not even have lowercase letters; the entire text was capitals run together.

Reading was hard work, and the marks that exist today are, almost without exception, optimizations that were invented to make it easier.

Aristophanes and the three dots

Around 200 BCE, a librarian at Alexandria named Aristophanes of Byzantium proposed a system for marking pauses in text. He used three dots, placed at different heights:

• A high dot for a long pause, like a modern period.
• A middle dot for a medium pause, like a modern colon.
• A low dot for a short pause, like a modern comma.

His system was rhetorical, not grammatical. The marks told the reader where to breathe and how long to hold the silence; they did not signal sentence structure as we now think of it.

This rhetorical theory of punctuation persisted for the next 1500 years. Marks were stage directions for reading aloud. The shift to grammatical punctuation — marks that show how phrases relate to one another, regardless of how they are spoken — happened only with the printing press, and even then unevenly.

The medieval invention of the question mark

The question mark is younger than the comma but its origins are more mysterious. The leading theory, traced by typographer Keith Houston, is that medieval scribes wrote the Latin word questio at the end of an interrogative sentence, then abbreviated it to qo, then stylized the q above the o until the resulting glyph became the modern question mark. There is no surviving manuscript that proves this evolution step by step — the trail is reconstructed from later forms — but the proposal is plausible and the alternative theories are weaker.

Note what this means: the most universal piece of modern punctuation, the one that has spread to almost every writing system on Earth, may have started as a single scribe's personal shorthand. Standards do not always come from authority. Sometimes they come from convenience repeated until everyone copies it.

Aldus Manutius and the comma

The modern comma, the one that hooks below the line, was not used by the Romans. The Latin word comma meant a phrase or short clause, not a mark on the page. The mark we now call a comma was popularized by Aldus Manutius, the Venetian printer who in the late 1400s set out to standardize Italian printed text.

Manutius was responsible for an astonishing amount of what we now consider basic typography. He invented italic type. He developed the modern semicolon for separating clauses too closely related for a period and too distinct for a comma. He pioneered the small portable book, the octavo, which transformed reading from a desk activity into something you could do on a journey. And he settled the shape of the comma — a small curved tail descending from the baseline — that we still use.

The semicolon is worth noting on its own. It was a deliberate Renaissance invention, a hybrid mark for a hybrid relationship. Most languages still use it, but few writers use it well. George Orwell tried to write an entire novel without one. The contemporary suspicion that the semicolon is precious or showy is, in a sense, accurate: it was invented to give writers more options, and any extra option will eventually be misused.

The space, briefly

Word spacing arrived in Europe through Irish monasteries in the seventh and eighth centuries. The monks were copying Latin texts in a language none of them spoke natively, and the unbroken stream of letters was unmanageable. They began inserting small spaces between words to make parsing easier.

The practice spread slowly, then all at once. By the eleventh century, spaces between words were universal in Western Europe. By the fifteenth century, with the spread of printing, they were inviolable. It is hard to imagine reading without them; it is harder to remember that ten centuries of European literature was written without them.

The space is, formally, a punctuation mark. It just happens to be invisible.

Quotation marks, the most contentious mark

Quotation marks have a less dignified history than the others. Their ancestor, the diple, was a small wedge that medieval scribes drew in the margin to mark important passages — usually scriptural quotations. By the printing press era, the diple had migrated to the inside of the line and was placed at the start and end of quoted speech.

What it looked like depended on the printer. The English convention used hooks above the line. The French convention used angled brackets, the guillemets. The German convention placed the opening mark below the line and the closing mark above. The Japanese convention used corner brackets that fit the grid of vertical text.

The result is that quotation marks are the most regionally divergent punctuation in widespread use. There is no global standard. Software still gets confused. The rules for whether other punctuation goes inside or outside the quote vary by country and by publication. American conventions place commas inside the quote regardless of logic; British conventions place them according to whether the comma is part of the quoted material. Both groups think the other is doing it wrong.

The marks that almost made it

For every mark that survived, several did not. The interrobang — a combined question mark and exclamation mark for sentences that are both — was invented in 1962 by an American advertising executive named Martin Speckter. It got typed onto Remington typewriter keys in 1968. It died of indifference; no one needed it badly enough to keep it alive.

The irony mark has been proposed at least four times in the last 200 years, by writers who wanted a way to flag sarcasm in text. Each proposal failed for the same reason: irony does not survive being labeled. The mark would defeat its own purpose.

The percontation point — a backwards question mark — was proposed in the late 1500s by Henry Denham as a way to mark rhetorical questions, the kind that do not expect an answer. It survived into the early 1600s and then vanished. We still need a way to distinguish rhetorical from genuine questions in writing, and we still do not have one.

The marks we use are not the marks we need. They are the marks that won.

What this changes

Punctuation feels like part of the natural world because it predates everyone alive. But every mark on the page is a tool that was invented for a purpose, argued over, refined, and standardized through a process that is still going on. Emoticons, then emoji, then code-switching between formal and texting punctuation, are continuations of the same old project: making written text carry more of the information that spoken language carries by default.

The marks will keep changing. The semicolon may finally die. The interrobang may yet have its day. And someone, somewhere, will need a punctuation mark that does not exist yet — and they will invent one, and if we find it useful enough, we will copy it until it is universal, and a thousand years from now no one will remember it had a beginning at all.