Written material from Greco-Roman antiquity survives in four forms: inscriptions, mostly on stone, but sometimes on wood or metal; papyri, made from the papyrus reed; metal coins; and books.
Of these four, the first three can be meaningfully grouped together, firstly as things actually made in antiquity, and secondly as largely accidental survivals. An inscription from 5th-century Athens is just that: a man carved it into stone in Athens 2,500 years ago. But in most cases, that inscription has only survived because it was buried before it could be destroyed or recycled. Likewise, most ancient coins survive from buried hordes, and papyri from the rubbish dumps of Egypt. So most of this material was once discarded, and indeed it is only as it has become the documentation of a distant world that is has grown in interest. A land rental contract is of no immediate fascination to those not affected by it, but a 2,000 year old land rental contract is a mystery worth unravelling.
The case is quite the opposite for ancient books. True, thousands of fragments from books, and a few complete works, have been recovered from binned or stored papyri, and there are scores of surviving parchment manuscripts from the last centuries of antiquity. But the vast majority of ancient books, in their earliest surviving form, are not ancient objects, but medieval ones. For most classical texts, there is no surviving copy earlier than then 9th century AD, and for quite a few, none earlier than the 15th.
The vast majority of ancient books, in their earliest surviving form, are not ancient objects, but medieval ones.
We have enough earlier survivals, and enough accounts of ancient book production, to know that these medieval books are very different sorts of objects from ancient ones. The ordinary form of an ancient book was a papyrus roll, written in what we now call capital letters (the only sort that then existed), with no spaces between the words. In some ancient fragments, words are separated by dots, but it seems the ancient were mostly used to reading without any sort of word separation (THISSEEMSCRAZYBUTONECANGETUSEDTOIT).
The medieval book, in contrast, is quite a familiar looking object. It is what scholars call a “codex”, i.e. a book bound between covers, that is read by turning pages, rather than a scroll of sheets stitched end on end, that one reads by unrolling. It is written in lower case letters, with capitals used for various forms of emphasis. Words are separated by spaces. Punctuation, while not yet standardized as we know it, is employed intensively. Many medieval books even have such familiar features as running titles at the top of pages and tables of contents.
Where medieval books—or “manuscripts”, as we have come to call them since the rise of printing—are in Latin, many look still more familiar, due to the form of their handwriting. In the 9th-century, under the aegis of Charlemagne, an extraordinarily clear new script was developed for book-copying, known today as “Caroline minuscule”. This script was replaced in later centuries by the angular Gothic scripts familiar from the fonts imitating them in some old printed books, notably in German. But Caroline minuscule was revived in the 15th century by Italian Renaissance humanists, who had rejected the increasingly messy Gothic scripts as barbaric. This revived Caroline minuscule was then the model for many of the early fonts used in printing. The script in which you are reading this article is closely related to Caroline minuscule, and you could read a Carolingian or an Renaissance manuscript without great difficulty.
Moreover, to think of the survival of an ancient text only in terms of its oldest extant copy is misleading. In most cases, that copy is only one of many—“many” can here be a few dozen, or a few hundred, or thousands. We do no think of Shakespeare as “surviving” in the extant copies of the octavos, quartos and folio editions printed in his lifetime. Those editions are of import to scholars of Shakespeare’s text, but, if they were all lost, Shakespeare would survive just as well as before: in truth, the “survival” of Shakespeare is constituted of all the millions of copies of his plays and poems, in original and translations, that have been produced from his time to our own.
The “survival” of Shakespeare is constituted of all the millions of copies of his plays and poems.
A book that forms part of a living literature is not a single object, be it handwritten or printed or indeed digital, but a chain of transmission, of which each of those objects is only a link. The book survives until every link in that chain is obliterated. Conversely, if even one link remains, the chain can always grow again.
Books may indeed sometimes survive by one single, haphazardly preserved link—this is the case even of some Latin and Greek texts. Likewise, the discovery of the Dead Sea Scrolls brought to light the library of a forgotten sect. But such resurrections are not the essential story of the survival of ancient books in Greek and Latin: even such famed “rediscoveries” of the Renaissance as Lucretius and Tacitus were in fact of medieval manuscripts in monastic libraries, not buried antiquities. In general, while the stream of ancient books shrunk here and there to a trickle, or flowed underground, it never ran dry, and spread into many a great lake, even when every copy had to be made by hand, and in a world where few could read.
Thus ancient texts have never been entirely ancient, never reduced to the relics and remnants of a discrete past. Latin and Greek were never forgotten languages, needing to be rediscovered and relearned, and ancient literature has again and again flourished as a modern literature, with new copies made for new readers.
Our relations with the past always consist of both continuity and rupture. But it is above all the continuity, however fragile at times, of writing and reading, from when Homer was first written down till today, that has given us the books of antiquity.
