The Politics of Punctuation: Changing History One Mark at a Time

Florence Hazrat

If you’re an atypical millennial or a boomer rejecting newfangled stuff, chances are you don’t how social media works. Fear not, the internet has you covered. Illustrator Jessica Hische kindly explains all about ‘following’ friends and enemies, snappy posts, and ‘sliding into DMs’ on the blue-birded micro-message platform on her self-referential website Chances are, it took you a bit to tell the words of the name apart, but you were able to read the domain name easily enough. On Twitter itself, we can make this a little easier through ‘camel case hashtagging’ (#LikeSoWhereEachNewWordIsCapitalised), or through ‘hyphen-strings’ ( Owing to complicated overlapping programming languages and the need for chunks of information to remain stuck together, neither website names nor hashtags can accommodate word spaces.

Squashing words together online can lead to fun and games. Consider one of the longest domain names in existence: it sandwiches the permissible number of 63 letters between the forward slash and dot com:

The site pokes fun at internet rules and regulations, offering nothing but a single blank landing page announcing ‘God’s final message’, a quotation by Douglas Adam: ‘We apologize for the inconvenience.’

Half analyst, half therapist (it’s pronounced “anALrapist”).

Distinctly more awkward are uncomfortable word neighbours like (a directory for therapists near you), or (all that a fountain pen lover ever dreamt of), or (not a gentlemen’s special interest website but one for the equally particular astrological pastime of documenting the position of the sun in relation to the earth).

Without latest technologies, it seems we have all but gone back to the roots: INANCIENTGREECEANDROMEWRITINGLOOKEDJUSTLIKEAMODERNWEBSITE WITHOUTSPACESBETWEENWORDSORANYMARKSOFPUNCTUATION. It seems we have come full circle. What are the steps between, though, and is this really the whole story of punctuation in antiquity? A dot, it turns out, is a powerful force for social change, however unassuming and meek it seems. So here is its story.

The Codex Sinaiticus, the most famous Greek manuscript of the Bible, written in the mid/late 4th cent. AD; its leaves are divided among four libraries, with the majority being in London’s British Library.

Helping Inexperienced Readers

In the beginning was the wedge pressed onto soft moist clay by a reed stylus somewhere in Mesopotamia 5,000 years ago. Cuneiform, the oldest writing script consisting of wedge-shaped marks on clay tablets, does not contain punctuation, or rather experts aren’t sure whether the small spaces between individual marks can be understood as punctuation or are merely a function of giving each sign its own airspace, and whether a certain wedge combination may mark a syntactic boundary.

Votive tablet to King Hammurabi featuring cuneiform, 1792–1750 BC (British Museum, London).

Ancient Egyptian hieroglyphs, slightly younger than their cuneiform siblings, show the first attested marks of punctuation, broadly speaking: writers manipulated the layout of pictures by drawing horizontal lines between paragraphs, started new lines to isolate discrete units of sense, ‘rubricated’ to draw attention to particular words like titles or names of deities, and included a pause sign.

Although not exactly widespread, both the concept and the systematic practice of punctuation did exist in pre-classical writing systems, and its twinned functions will become familiar throughout the history of punctuation: it facilitated comprehension, including live recitation, and it aided those learning to read by creating boundaries where one sense stops and another starts.

Hieroglyphs from the Stele of Minnakht, chief of the scribes; produced during the reign of Ay, c.1321 BC (Musée du Louvre, Paris, France).

We remain in Egypt, but fast forward a couple of centuries BC, to a time when Greek civilization dominates the Mediterranean basin, but mixes provocatively with pharaonic culture in Egypt. Trade, navigation, diverse populations with their own expertise and habits jostle and combine in Hellenic Egypt, making it an exciting and inevitably progressive place to be.

Around 300 BC, when the famous library of Alexandria was founded, punctuation conventions mainly negotiated the blank space of the papyrus page (or other writing surfaces) through the placement of the words themselves – that is, the grouping together of units of sense: if there had been an embryonic notion of paragraphs in earlier hieroglyphs, that understanding developed into a recognizable practice of sectioning off new topics, speakers, or stanzas either with a horizontal line, or a hook placed on the line but in the outer margin. Mostly, however, Greek text came about as a continual string of words, potentially going on forever (or until the scroll carrying the words came to an end).

A fragment from Book 8 of Plato’s Laws, written in the 3rd cent. AD and discovered in 1897 in the Egyptian city of Oxyrhynchus (P. Oxy. 23; Sackler Art, Archaeology and Ancient World Library (!), Oxford, UK).

It’s not entirely clear why scribes would squeeze words and sentences together without borders. Perhaps writing was not considered a separate manifestation of speech, but rather a mere record, and since we don’t pause between words when we talk (or we do, but imperceptibly), it didn’t occur to people to mark word boundaries through a mark or space. Such radical punctuation developments would take another thousand years, and happen on a far-flung island at the edge of Europe – 8th-century Ireland. But that’s another story. The task of any research library, however, is to preserve and transmit knowledge for future generations, and for those who precisely do not know. So thought Aristophanes of Byzantium (c. 257–180 BC), the third in the illustrious line of Head Librarians of Alexandria, all of whom made a name for themselves in punctuation.

Greek magical papyrus of the 3rd cent. AD (Papyri Graecae Magicae 121 = British Library P.Lond. I 121).

In order to help non-native Greek speakers access texts and pronounce the language properly, Aristophanes suggested putting a dot at the top level of the line to mark a ‘period’, or full sentence; a dot at the bottom of the line to mark the end of a ‘colon’; and a dot at the centre of the line to divide two ‘commas’ from one another. The terms did not at first refer to the actual signs we now use (, and : and .) but to the grammatical units between them. Apart from signalling grammatical boundaries, a ‘period’ dot necessitated a long pause, a ‘colon’ dot a medium pause, and a ‘comma’ dot a short pause.

But why dots? A dot is a genius idea. Whatever your writing material, a dot is always going to look the same. One would be hard-pressed to mistake a dot for any other Greek letter. Whether your handwriting is sloppy or symmetrical, a dot is going to look the same. It’s that indivisible focal point. In his whimsical essay ‘The Philosophy of Punctuation’, cultural critic and thinker Paul Robinson declares “the powerful circular shape of a bullet is noncommittal to the area it occupies and is unrelated to the type-style and spacing.”

A dot is also a smart choice, because it’s tiny enough to insert between any two letters at any time in the process of writing (or indeed afterwards). So, if you forgot to punctuate, or if you want to punctuate a text long after it’s been copied, you can just go ahead and do that right into the manuscript. The dot is an unassuming sign, but a momentous invention. Perhaps momentous precisely because it doesn’t brag. It’s just… there. Full stop.

A speech delivered by Emperor Claudius, surviving in a cursive manuscript of AD 41–54 (Berlin Staatliche Museen BGU 611 = P. 8507 R).

When in Rome

Whether Aristophanes’ system did catch on widely is not entirely certain: existing papyrus fragments are few and far between, but knowledge of the system survived thanks to the second-century BC grammar of Dionysius Thrax (170–90 BC) which became a staple of Classical education for centuries. The Romans knew of Dionysius’ grammar too, but didn’t bother to take on Aristophanes’ system straight away. They just continued using the old script, with no spaces or signs at all. The dot, usefully low-maintenance as it is, does stick around, however. Just not in manuscripts.

Adding a blank between words or lines of text is the easiest thing when writing on paper or Microsoft Word. You tap the space bar, and that’s it. Things look a bit different online where one needs to know some tricks of the trade in order to leave that kind of space. Instagram and Facebook mush all your nicely parcelled-off paragraphs together unless you put a full stop as single-line space-holder. As with the concept of scrolling, this practice goes way back. Before spaces between words was the norm, words or sentences could be divided by dots, or ‘interpuncts’. The earliest known example is the Mesha Stele, a big black stone whose inscription in an old Hebrew script dates back to 840 BC, describing the exploits of King Mesha of Moab. Every word is set off from its neighbours by dots hovering around the centre of the line. It’s easy: hold the chisel to the rock, hit its handle with a heavy object, and there is your interpunct.

The Mesha Stele, a Canaanite inscription of 840 BC (discovered in Dhiban, Jordan, in 1868 and now in the Musée du Louvre, Paris, France).

Interpuncts were truly international: while Aristophanes was busy inventing his dot system, Etruscans and early Romans divided their stone-carved words with dots. Interpuncts would, however, remain an interlude in the history of Roman punctuation.

Like the Greeks, Romans believed that writing was merely ancillary to speech: they valued the ability to hold speeches and convince through words, as their education was primarily directed towards teaching oratorical skills. Thinking happened in sound, not sight. But what if you weren’t sure about sound? Consider this: COLLECTAMEXILIOPUBEM. Is that COLLECTAM EX ILIO PUBEM (“a people gathered from Troy”) or COLLECTAM EXILIO PUBEM (“a people gathered for/from exile”)?

We don’t need to understand Latin to see the potential for diverging interpretations, and the ambiguities that result without word spaces. In fact, the Romans themselves weren’t sure about this one: the prominent grammarians Servius (4th century AD) and Donatus (late 4th/early 5th century AD) disagreed over the syllable boundaries of this line. It’s from Book 2, line 798 of the Aeneid, the Latin epic that Virgil (70–19 BC) composed between 29 and 19 BC, which affirmed the founding myth of the Roman empire as the successor of the great fallen city of Troy (also known as Ilium).

Aeneas flees From Troy, Lucca Batoni Pompeo (attr.), 1754–7 (Galeria Sabauda, Turin, Italy).

The Trojans kept Helen of Troy throughout a decade-long siege of their city by the Greeks; Book 2 of the Aeneid recounts how Ilium burned down only after a majority of Trojans were deceived by Odysseus’ famous ruse with the Trojan Horse. Most Trojans died, but some escaped, including the hero Aeneas, son of Anchises and the goddess Aphrodite. With his aged father on his back, and his little son held by the hand, he fled the smoking ruins with other survivors and sailed to Italy, where his descendants would eventually found Rome.

So, did Aeneas “gather a people from Troy/Ilium” (ex Ilio) or did he “gather a people for exile” (exilio)? Both options remain attractively open, both in terms of sense and the rhythm of hexameter verse, and perhaps Virgil intended it so.

Romans were well aware of the ambiguity provoked by continual, unbroken script (scriptio continua), but punctuation (of whatever form) would mean imposing a single reading on the text, interpreting, and thus demystifying, a deliciously layered literary work like the Aeneid. Punctuation could also involve the danger of claiming a particular ‘correct’ interpretation, a task which involved literary judgment, a task not often associated with the professional scribe or secretary.  Although often highly accomplished, these men were mostly freed slaves, and their role was expected to involve an uncritical attitude towards the text they copied. They were mere replicators, reproducing identical texts, but not interpreting them.

A fine example of Roman capitals: the Lyon Tablet, a bronze inscription of a speech by Emperor Claudius, c. AD 48 (Gallo-Roman Musem, Lyon, France).

Punctuation unpicks ambiguity in writing by creating boundaries between words and sentences, and by suggesting relationships between different clauses. It helps fix meaning (though by no means always, and it’s often those very glitches which become exciting). As such, punctuation was of limited use to the Romans, who, at times, revelled in literary ambiguity

Reading (or giving a speech based on your reading) required assiduous preparation: orators would devise their individual system of marks signalling which syllables belonged together and where to pause for emphasis or sense. In Ancient Rome, there was no such thing as reading at first sight. The fear to bungle a text was real: Aulus Gellius (AD c. 125–80) refused to read a paper publicly that was shoved into his hands, saying “How can I read what I do not understand? What I shall read will be confused and not properly phrased.” Someone in the crowd pushed himself forward, intending to show off, but he “cut up the sense and pronounced the words wrongly” (Noctes Atticae, 13.31.5).

Those few who were capable of scanning the letters with their eyes, and immediately grasping their relationships as they were pronouncing them, were held in high esteem: Trimalchio, a grossly vulgar character in Petronius’ opulently absurd work Satyricon (60s AD), passionately kisses a boy, but, when confronted with the act by his wife, alleges it wasn’t for the boy’s beauty, but because “he can sight-read a book” (librum ab oculis legit).

Roman relief of three pupils and a teacher, AD 180s, from Neumagen near Trier (Rheinisches Landesmuseum, Trier, Germany; photograph of cast in Pushkin Museum, Moscow, Russia).

The boy’s wonderful skill (if he did have it) should make sense – at least, that’s what the masters of Latin oratory and persuasive speech Quintilian (AD c.35–100) and Cicero (106–43 BC) thought. In their rhetorical manuals and other observations on writing and speaking, both rhetorical superstars defend the position that punctuation is unnecessary. If, that is, the writing is good. And the writing will be good if the rhythm is good, and the rhythm is good if you are a capable grammarian who knows how to put well-sounding words together. It ought to be the rhythm of words, and their units, that tells you about their boundaries. In De Oratore, for example, Cicero described how a sentence by force ends because of certain features inherent in eloquent Latin rather than because of the signs introduced by a scribe: readers “wanted clauses in their speeches punctuated with marks of where to breathe, not because of our tiredness or that of the scribes, but according to the words and sentences” (interspirationis enim, non defatigationis nostrae neque librariorum notis sed verborum et sententiarum modo interpunctas clausulas in orationibus esse voluerunt, De Oratore 3.173).

Rhythms, Roman orators argued, basically constrain the speaker into a specific breathing hence pausing pattern, rendering punctuation superfluous. Familiarity and instant grasping come with practice and experience. Traffic signs dividing words from one another gave away the learner, not the master: in his rhetorical textbook Institutio Oratoria, Quintilian mentions marks added into text for schoolboys, who when practising reading out loud gaveemphasis to particular words through their voice and posture: “[reading remains to be discussed] so that the boy may know where he should suspend his breath, and where to distinguish the verse, where the sense ends” (puer ut sciat ubi suspendere spiritum debeat, quo loco versum distinguere, ubi cludatur sensus, 1.8.1). Punctuation, the damning assumption goes, is for babies.

Young man reading a scroll, fresco from Herculaneum, 1st cent. AD) (National Archaeological Museum, Naples, Italy).

Punctuation as Semaphore

Although Cicero mocked those who needed copyists’ strokes to navigate the high waters of his sea of letters, evidence in manuscripts shows that not everyone was a Latin literary giant. Just as in Ancient and Hellenic Egypt, punctuation in Rome served learners to train themselves in getting to grips with the grammatical ins and outs of the language. Teachers of grammar would pre-punctuate texts in a learning stage known as praelectio, or ‘pre-reading’, which might include apices (signs designating how long a vowel had to be pronounced), interpuncts to signal word boundaries, as well as hyphens and ‘diastoles’ – hyphens were below-the-line bridges between syllables connecting tricky words that belong together (just like hyphens do today); diastoles, on the other hand, disconnected words which had erroneously run together, like how ‘an adder’ arose from misconstruing ‘a nadder’  (from nǣddre, the Old English for “snake”) as ‘an adder’.  There was also a 7-shaped mark called simplex ductus slipped in between one general matter and another, while an oblique stroke like this | would mark a minor pause after a colon or comma, signposting the structure of the sentences as a whole.

Scribes or students who copied texts anew could also work with the layout on the page, distributing verse texts line by line, indenting new paragraphs, or leaving slight spaces after a period ended. The paragraph sign as we know it today (¶) as well as the hedera, or ivy leaf (❦), would separate sections within a text, blending decorative features with markers of sense boundaries, and hence breathing pauses. The beginning of a sentence also offered opportunities to clarify the structure: sometimes, a ‘K’ for kaput or ‘head’ would stand at the beginning of a sentence, drawing attention to new subject matter (hence our modern ‘chapter’); sometimes enlarged first letters (so-called litterae notabiliores, or‘notable letters’) would be set off in the left-hand margin of every paragraph, although it would be another thousand years until the concept of upper and lower case letters would circulate widely across manuscript culture. Until then, a set of marks (some of which were systematized, some of which remained idiosyncratic), along with the manipulation of space on the page, would be used to help helpless readers understand and navigate a text.

Although this manuscript by Sulpicius Severus on the life of Saint Martin of Tours from AD 517 is too late to represent Roman practice, it does carry over some older features. It leaves spaces between sentences, and has decorative lines between paragraphs (Verona, Biblioteca capitolare, MS xxviii (36), f.31).
Note the ivy leaves in lines 2 and 3, marking two sections off each other in this 6th-cent. manuscript of Pope Gregory the Great’s Cura pastoralis (Troyes, Bibliothèque municipale, MS 504).

This kind of punctuation was not just for reading’s sake but also for literary interpretation: notae, or marks singling out whole passages for special attention, were derived from those first Alexandrian librarians and their efforts to establish readable and most of all correct texts in the vaults of their vast library. The first Head Librarian Zenodotus (early 3rd century BC), when trying to establish an authoritative text of Homer, drew a straight line in the margin for any passage of the Iliad or Odyssey that he judged to have been wrongly attributed to the poet. Zenodotus was followed by Aristarchus of Samothrace (c.220–143 BC) who put an asterisk (*) in the margin for any genuinely Homeric lines of verse that were duplicated elsewhere in the long epic, and a ‘diple’ (>) to mark his disagreements with Zenodotus’ previous choices.

The ancient asterisk (top left) and diple (bottom right), from an 8th-century copy of Isidore of Seville’s Libri etymologiarum (Wolfenbüttel, Herzog August Bibliothek, Weissenburg MS 64, f.13).

By the time of the Romans, the diple’s function had migrated to drawing attention to something noteworthy in and of itself. A couple of hundred years later, in the 3rd century AD, the theologian and church father Origen (AD 185–253) edited the Septuagint, that is, the Greek translation of the Old Testament.

Like a true scholar, Origen gathered together several versions, and compared them, trying to establish a critical edition, complete with those Alexandrian notae which tell readers about doubtful or re-shuffled passages. Anytime Origen quotes from the Bible in his edition-cum-commentary, he uses the diple (since it’s a noteworthy passage, being the word of God), and so the diple became associated with quoting – an announcement that these are the words of someone else, from somewhere else. Perhaps that’s why continental quotation marks come as double diples, or «guillemets», rather than English “stalactites”, suspended from the top of the line.

By Late Antiquity, praelectio signs as well as Aristophanes’ dots had entered manuscript culture, becoming codified by Emperor Charlemagne’s inspired education reform adviser, Alcuin of York (735–804). Our modern marks are their heirs.

Alcuin (middle) attends his fellow scholar Hrabanus Maurus (left) as he presents his work to the dedicatee Archbishop Odgar of Mainz (right), miniature of c.831 (Vienna, Österreichische Nationalbibliothek, cod.652, f.2v).

Punctuation as Subversive Force

Latin readers and writers had the option not to punctuate at all, or to do things with their writing in order to navigate it more easily. Like textual semaphore on the blank sea of the page, paragraph signs, ivy leaves, dots, and differently-sized letters advertised relationships between portions of words, how they relate to one another, and how to synchronise breath with thought. This orchestration of meaning gave access to texts, and thus to knowledge and information, to a broader range of readers, including non-native speakers and those with fewer years of education.

Punctuation became the key to unlock the mystery of the page, making it both easier to read, and faster. Information circulated more quickly around a greater number of people, enabling increased communication and speed for all sorts of human activities, from scholarship to trade. Then, as now, information is power, so enlarging the pool of people who gain it meant softening up rigid social structures.

While punctuation certainly isn’t the most significant of human inventions, it is a force to be reckoned with regarding literacy and its concomitant benefits. After thousands of years of marking text, it’s legitimate to ask quo vadis, punctuation? Where are you heading? In comparison to Aristophanes and friends, our era’s contribution to disambiguating written meaning seems distinctly underwhelming: yet again, we are ceaselessly borne back into the past, returning not to continual script but even further back to pictograms. For are not emojis the hieroglyphs of the 21st century? The real question is whether the billions of hearts and smiley faces we send one another every day still obtain some of that radical social impact which their drab comma and colon cousins possess.

Florence Hazrat is a Berlin-based independent scholar of Renaissance literature and a writer on language and literature for a broad audience. She has written a whimsical biography of the exclamation mark called An Admirable Point (Profile, London, 2022). 

The image at the head of this article is Eustathius’ copy of his commentary on Homer’s Iliad, written in the 12th century (Florence, Biblioteca Medicea Laurenziana, Plut. 59.2, f.1r).

Further Reading

John Bodel, “Paragrams, punctuation, and system in Ancient Roman Script,” in S.  Houston (ed.), The Shape of Script (SAR Press, Sante Fe, NM 2012) 65–92.

Keith Houston, Shady Characters: The Secret Life of Punctuation, Symbols & Other Typographical Characters (Norton, New York, 2014).

M.B. Parkes, Pause and Effect: An Introduction to the History of Punctuation in the West (Scolar Press, Aldershot, 1992).

Paul Robinson, “The philosophy of punctuation,” in Opera, Sex, and Other Vital Matters (Chicago UP, 2002) 303–7.