What is a paragraph? Consult a writing guide, and you will receive an answer like this: “A paragraph is a group of sentences that develops one central idea.” However solid such a definition appears on the page, it quickly melts in the heat of live instruction, as any writing teacher will tell you. Faced with the task of assembling their own paragraphs, students find nearly every word in the formula problematic. How many sentences belong in the “group?” Somewhere along the way, many were taught that five or six will do. But then out there in the world, they have seen (or heard rumors of) bulkier and slimmer specimens, some spilling over pages, some consisting of a single sentence. And how does one go about “developing” a central idea? Is there a magic number of subpoints or citations? Most problematic of all is the notion of the main “idea” itself. What qualifies? Facts? Propositions? Your ideas? Someone else’s?
In his 1928 English Prose Style, the poet and art critic Herbert Read argued that there’s no point in fussing about the “vague” notion of a central “idea” anyway, since it “will be found of little application to the paragraphs we find in literature,” a claim that Read illustrates with unruly precedents from Thomas Babington Macaulay, John Milton, and D.H. Lawrence, among others. What Read clarifies is not only that single-minded definitions buckle under even minimal stress. Taking up his nearly century-old book, one recognizes a peculiar tradition in which one textbook after another, one generation after another, has promoted a blueprint for paragraph construction conspicuously at odds with the prose of the most highly acclaimed stylists of the English language.
What gives? The tension reflects the paragraph’s curious history as a punctuation mark and unit of thought. In fact, my opening question—what is a paragraph?—only gets more complicated as we gaze further and further into the past, as the paragraph gradually dwindles to a thin line in the margins. This backstory explains why it is so hard to say what exactly a paragraph is and, in turn, why we struggle now to legislate its parameters. But this isn’t an entirely despairing story: To recall the paragraph’s past lives is also to consider how previous generations have put their thoughts in order and to gain thereby a vantage to reconsider our own writing practices.
The trouble begins with the ancient Greeks. Their scribes—and later their Roman imitators—laid out documents in columns on papyrus bookrolls (a.k.a. scrolls) using a method known as scriptio continua in which words are written without spaces in between. The classicist William Johnson has memorably likened the effect to “a tight phalanx of clear, distinct letters, each marching one after the other to form an impression of continuous flow.” But scriptio continua poses an obvious challenge: The reader must sort the marching characters into meaningful words and sentences. Unsurprisingly, scribes and readers over the centuries invented marking systems to aid the reader’s labors of understanding and, equally important, vivid articulation—reading being very much an oral performance in antiquity.
The first such mark—in use from the fourth-century BCE on—was a plain horizontal stroke drawn in the margin alongside or perhaps slightly intruding between lines of the text. This paragraphos (literally, “written beside”) has been called “the first punctuation mark,” though it likely wouldn’t pass muster with modern grade school teachers because it didn’t have a consistent grammatical or rhetorical function. It signified simply that a transition of some kind would take place in the neighboring line—perhaps the beginning of a new sentence or stanza, perhaps a change of speaker in a drama or Platonic dialogue. Typophile Keith Houston has rightly called the paragraphos a “crude instrument.” Its pliability, though, made it eminently useful.
As a mark of change, the paragraphos was a familiar device in the scribal arsenal—along with techniques such as outdenting, enlarging letters, and leaving empty space— for identifying subsections of texts, including those that conform to our sense of paragraph-scale. However, and here we run into our first bump in the narrative, classicists and biblical scholars have debated whether to call these chunks “paragraphs,” at least in the modern sense. First of all, save for a few hints otherwise, these marks cannot be attributed to the authors of the documents; they represent a later (perhaps centuries-later) reader’s or scribe’s interpretation of a given document’s structure (which sometimes varies between copies). More importantly, classical rhetoric had no concept of “the paragraph” as “a generic unit of discourse,” as the rhetorician Jeanne Fahnestock has observed. To be sure, ancient rhetoricians were formidable scholars, and left behind an enormous body of useful counsel about language (poetic and prosaic), argumentation, and education, among other matters. But their principal charge was the training of orators, and though some teachers encouraged writing exercises to that end, none taught the skill of assembling a series of written blocks of text, each designed to unfold ideas, themes, subjects, incidents, etc. Antiquity, in short, provided the terminology from which the paragraph derives but no edicts to govern its production.
The medievals gradually disbanded the scriptio continua phalanx. First its field was taken. In the late Roman Empire, the bookroll was displaced by the new stack-and-flip writing technology, the codex (what we usually mean by “book” now), which had been adopted early on by Christian communities and was better suited to northern lands where papyrus was hard to come by but animal skins weren’t. The codex introduced the page—a new surface, framing device, and interface whose possibilities scribes and artists of the High Middle Ages would consciously exploit. But the more immediate threat to scriptio continua was the difficulty it posed for young monks with dodgy Latin. Transitional figures, including St. Jerome (c. 342–420) and Isidore of Seville (560–636), made efforts to make words and clauses more easily discernible to novice readers by recycling old practices of dotting and aligning clauses to increase readability. The more decisive change, though, was the development of what the paleographer M.B. Parkes dubbed a new “grammar of legibility” in the eighth and ninth centuries—among Irish monks, their English counterparts, and, further still, within Charlemagne’s realm (the so-called “Carolingian Renaissance”). These scribes employed multiple strategies to produce a more legible text, including devising the first miniscule script (i.e., “lower-case”), but none was more conspicuous, or momentous, than their practice of leaving space between words as they wrote. The book historian Paul Saenger likened an early form of this style of writing to aeration—as if the words were at last given room to breathe on their own.
Like their ancient predecessors, medieval scribes had a number of ways of identifying subsections, including the paragraphus (note the Latinization), albeit now in its mature classical form resembling a Greek gamma. (Isidore was still, influentially, commending it as a strategy for signaling beginnings in the seventh century.) Yet in the High Middle Ages, the paragraphus would be eclipsed by a mark known alternatively as a paraph or capitulum, whose origins lie in the Roman practice of placing “K” (from the Etruscan word for “head,” kaput) at the beginning of chapters and sections. Over the centuries, that K became a C, and the C was later merged with slanting bars, ultimately producing the modern paragraph mark, known in English as the pilcrow (¶).