Codex

The papyrus scroll was eventually displaced by the parchment codex, seemingly a Roman invention, and it is at Rome that epic in codex form makes its first appearance. The Saturnalia gifts epigrammed by Martial towards the end of the first century ce include books: heading the list, Homer’s Battle of Frogs and Mice, book-form unspecified and hence to be imagined as a papyrus scroll; but others are spectacularly miniaturized parchments (membranae) - Homer, Virgil, and Ovid’s Metamorphoses among them. More than a century earlier Cicero recorded a parchment Iliad enclosed in a nut-shell, a still more remarkable feat of miniaturization; that is thought more likely to have been in scroll form, but Martial’s Homer (Iliad and Odyssey in one) is expressly in pugillaribus membranis, and that unequivocally implies codex form. Pugillares (‘‘fistfuls’’) had hitherto been wooden notebooks. The parchment codex in fact represents the fusion of two separate media, and proceeds more from the reapplication of pre-existing technologies than from the invention of new ones. The original codex (or caudex), composed of wooden boards, has little place in an account of the media of epic; the diptych and related forms were looked at above in the context of cuneiform literature, but in the Greco-Roman world the literary use of such tablets is confined to informal verse, first drafts, or school exercises.

Parchment is another matter. Skins (Gk. diphtherai, Lat. membranae), as we also saw above, were in use in western Asia around the middle of the first millennium and no doubt earlier, and Cicero’s Iliad in a nutshell, however much of a freak (and anecdotally reported), is a significant item of evidence for the literary use of parchment prior to the end of the Roman Republic, when (as still in Martial’s day) books were ordinarily papyrus scrolls. A strong line of continuity from Asia to Rome is provided by Pergamum, which gave its name to parchment and which Roman tradition associated with its invention. Development of parchment at Pergamum, perhaps spurred by disruption of Egyptian supplies of papyrus during the internecine turmoils of the Hellenistic kingdoms (R. Johnson 1970), will have led to its introduction to Rome in the earlier second century BCE; maybe Crates brought parchment along with Homeric scholarship. The processing of animal skins is both complex and varied (Reed 1972), and the lines of development from the earlier diphtherai to parchment scrolls, whether in technology or in usage, are impossible to chart or even to define; the distinction between leather and parchment is inadequate, and there is little material evidence before the Dead Sea Scrolls. It is quite conceivable that the Pergamene Library, in contrast with the Alexandrian, held not only papyrus-scroll books but also parchment-scroll ones; here if anywhere there is reason to query the assumption that because Alexandria set the bibliographic norms books in Egypt may be taken as representative of books elsewhere. But the Roman evidence suggests that parchment did not come into regular use for (non-biblical) literary texts until the new codex form had established itself.

Martial’s parchment miniatures were ahead of their time (a proposition inviting deconstruction, however). The papyrus scroll continues to be the proper medium for highbrow literature until at least the end of the second century, while the codex upgrades its respectability. The intersections are multiple, the dynamics complex: parchment challenges papyrus, codex challenges scroll. The switch from papyrus to parchment seems at Rome to have accompanied the primary transition of scroll to codex, but in Egypt (papyrus-land) was very gradual, and of limited success: even in late antiquity a Homer codex is still much more likely to be of papyrus than of parchment. The more significant development - for in the end it does not make so very much difference whether a book is made of papyrus or of parchment, or of paper - is the switch from scroll to codex, the form of book we still use today. It constitutes a radical change in the reading experience; and as a textual vehicle the codex has certain advantages. What in time proved to be the most important of these is precisely the distinction claimed by Martial for his Saturnalia curiosities: compendiousness. Quam brevis immensum cepit membrana Maronem: ‘‘How short a parchment has captured boundless Virgil!’’ A codex could conveniently accommodate vastly more than a scroll. In this sense the codex realized its potential at its inception, a most unusual phenomenon in the history of a new medium. But how large a part this practical consideration played in determining the codex’s early trajectory is hard to tell. Christian predilection for the codex markedly antedates widespread adoption of it for pagan literature (epic is thus shifted from center-stage), suggesting socio-religious factors over against purely pragmatic ones; and some early codexes do not seem to have contained more text than contemporary scrolls. But some certainly do, and recognition of the codex’s capaciousness must go far towards accounting for the unstoppability of the upstart’s success in overcoming the resistance of the traditional medium. Other advantages will have become apparent too. With a codex it is easy to flip through and find a particular passage; in one third-to-fourth-century Odyssey codex (it contained the entire poem, something no papyrus scroll ever did) such reference is facilitated by the relevant book-number being repeated at the top of each right-hand page. In fact the practical advantages lie wholly with the codex, at least from a user’s point of view, and the scroll had little with which to withstand the new form except its deep entrenchment - potent, but not potent enough. The scroll’s displacement is far from instantaneous. The codex does not overtake it until around the end of the third century ce (at any rate in Egypt: perhaps a little sooner at Rome and elsewhere?), and Homer scrolls were still being produced in the fifth, and probably beyond.

A codex consists of a number of folded sheets, stitched together at the fold. By virtue of the fold each sheet has two leaves (folia, sing. folium); each leaf has two pages, front and back, a. k.a. recto and verso (recto the right-hand side, verso the side you reach when you turn the page; the terminology presupposes left-to-right progression). A simple four-sheet (8-leaf, 16-page) codex could be constructed in either of two ways: (1) fold each sheet, then place the folded sheets on top of each other; or (2) place the (unfolded) sheets on top of each other, then fold; you then have a four-sheet ‘‘quire’’ or gathering. Method 1 replicates the construction of wooden tablet sets, but it was method 2 that dominated: a simple enough assembly method, but attended by two practical problems. One is physical: with a large number of sheets, the construction is put under strain, and the book becomes awkward to use. The solution lay in the multiple-quire codex, with quires of around four or five sheets apiece. But standardization took centuries, and most early codexes are composed of a single quire, even when the number of sheets is large. A fourth-century papyrus codex of Iliad 11-16 (!?) consisted of a single gathering of 31 sheets (62 folia, 124 pages), and there are still larger single-quire Christian codexes of the third century. The other problem concerns the organization of the text. In a quire the sequence of leaves will be consecutive only in the case of the innermost sheet; in a four-sheet quire (a quaternio), the innermost sheet will be pp. 7-10 (fol. 4r to fol. 5v), while the outermost one will be pp. 1-2 and 15-16 (fol. 1rv and fol. 8rv). Whether or not the writing of the text is done before the codex is stitched together, there is an appreciable amount of preplanning required to achieve the correct match between the text and the pages; this is another feature that gives the codex affinity with the clay tablet, over against the simplicity of the papyrus scroll, but the complexities of the codex are that much greater. In medieval scriptoria the whole process was thoroughly systematized, but in the early centuries control of the medium can be less than fully assured. A third-century papyrus codex of Iliad 1-6 (omitting the ‘‘Catalogue of Ships’’ from Book 2) was for some reason written only on the recto pages. By the fourth century the standard conventions are securely in place, and technically wondrous products such as the great biblical codexes are created.

Tradition and simplicity, and perhaps cost, sustained the viability of the papyrus scroll in the face of the codex’s annexation of its territory. At the same time, and not without success, papyrus adapted itself to codex form. A papyrus codex’s sheets were cut from ready-made papyrus rolls. (Thus the leaves of a papyrus codex may have sheet-joins, since its sheets do not coincide with the original sheets of the roll.) Though there is some debate about priority, the papyrus codex is clearly to be recognized as genetically secondary to the parchment codex, and papyrus is in fact fundamentally ill-suited to codex use: it does not take at all well to folding. Nonetheless, in this deracinated form papyrus continued to maintain a strong hold on books of the higher literature, at any rate in Egypt; in other areas no doubt the parchment codex dominated earlier, or from the outset, eventually to be challenged in turn by paper, introduced to Byzantium before the twelth century. The parchment scroll, for its part, mimicking the conventions of the papyrus scroll, had (except in Jewish religious contexts) more rapidly given way to the codex.

From a technological point of view, the parchment codex does not represent much of a breakthrough; it is hardly comparable with the invention of paper, for instance, or of movable type. Nor does it quite have the significance sometimes claimed for it in the survival of Greek and Latin literature. Texts were intermittently transferred from scroll to codex (and then from codex to codex), as earlier they had been transferred from scroll to scroll. Codexes held more than scrolls, to be sure, but transference from scroll to codex was certainly no guarantee of successful passage through to the Middle Ages, as Menander and other non-survivors can attest. In that process sheer luck was a large factor but not the only one: the surviving epics are those which were most favored throughout their history. None of this detracts from the fact that the codex’s ousting of the venerable papyrus scroll is a momentous occurrence in the history of media.

The codex redefined the nature of the book and transformed the experience of reading, scarcely less than the computer screen and hypertext are in the process of doing in our own day. For all that, the new form of vehicle made little impact upon epic itself, too grand and stolid a genre to be seriously impinged on by its medium. Whether the Flavian epics (see Chapters 37, 38, 36, by Dominik, Marks, and Zissos respectively) first circulated on scrolls or in codexes hardly matters; the poets did not write for one or the other form, they wrote epics, whose conventions, numbered ‘‘books’’ and all, were securely in place. But it was the codex that enabled Virgil to be a single tangible corpus (complete with appendix), and it was the codex that allowed the material of Homeric commentaries to be lassoed into the margins of the mother text. One does not have to be a technological determinist to see the codex’s capaciousness as taking on an active role in conditioning the reception of the transmitted poems. While the paradigm of the papyrus scroll carried over into the new form (sometimes to extremes: was no absurdity perceived in the four-column-per-page layout of the codex Sinaiticus?), the mold-breaking role of the codex is clear enough. The self-sufficiency of the poetic text is increasingly given up, as the classic austerity of the papyrus scroll is pushed aside by more colorful and variegated styles of presentation. Martial’s Saturnalia Virgil had anticipated modern practice in including an image of the poet’s face (advertised in the second of the two lines of the attendant epigram), and ‘‘coffee-table’’ books of Homer and of Virgil from later antiquity feature illustrations of the action. More generally, with the codex we get numbered pages, enlarged and decorated initials, all sorts of prefatory and ancillary material. Not all these features originate in the codex, but the coming of the codex heralds a sea-change in exploitation of potentialities. Differences of function and aspiration display themselves much more prominently in the codex than they ever did in the scroll, and repackaging options are taken advantage of. The book itself can now be a prestige object, a thing of tremendous beauty and value, the codex lending itself to far greater mediumistic variation than the papyrus scroll: pocket-sized or too heavy to lift. The emperor Maximinus’ Homer is said to have been a codex of purple vellum written in letters of gold; Alexander’s had been a papyrus scroll with no claim to distinction except that it was corrected by Aristotle.

In the precarious passage from antiquity to the Middle Ages the invention of minuscule script and the revival of Hellenism in ninth-century Constantinople mark a fresh beginning of sorts, as old uncial codexes were replaced by new minuscule ones. On the Latin side there was greater continuity, however tenuous: the Carolingian revival involved less change in the medium, and codexes produced in the fourth and fifth centuries, some sumptuous Virgils among them, survived to find a haven in European libraries. Such texts might be pedigreed by their subscriptions; signing and dating explicitly confers individual identity on a manuscript. Paper came into increasing use as a cheap but inferior alternative to parchment, and the later decades of the fifteenth century saw the arrival of print, a mechanism finally enabling genuine textual identicality in multiple copies. The new technologies left the medium otherwise intact: early printed codexes are simulacra of handwritten ones. The codex remained unchallenged until the advent of electronic media.