Cuneiform originates in clay, and the two make naturally good partners. Already well established in the third millennium not only in Sumer, where it is generally reckoned to have originated, but in Elam to the east and at Ebla to the west, the clay tablet was a textual medium shared by many different languages and cultures. The materials, unlike the scribal skills needed to use them, were not hard to acquire. Rivers provided the clay, marshes the reeds from which styli were cut. The production process was technologically simple. The soft clay was shaped and smoothed, the ‘‘wedges’’ (or ‘‘nails’’) that comprised the characters were imprinted with the stylus, to form those mysterious signs that somehow represented language: ‘‘The Lord of Aratta scrutinized the clay. But the spoken word was a nail’’ (Enmerkar and the Lord of Aratta). Whether or not clay merits the title of‘‘the best writing material yet devised by man’’ by virtue of its unmatched cheapness and durability (Oppenheim 1964: 229), the nature of the material goes a long way towards explaining the range and persistence of cuneiform writing systems in the Near East.
With papyrus scrolls, manufacture of the writing material does not need to be coordinated with the actual writing, but the production of written clay tablets is a single integrated process, which must have been thoroughly organized in the temple and palace centers. Clay quality, geographically variable, will have been a matter of consistency of composition, plasticity, both wet and dry strength, possibly color. The finished product, the written tablet, was air-dried; it could be oven-baked, but it seems fewer tablets were accorded this treatment than was once thought, for many of today’s tablets were baked only in fires of conquest or accident, or upon excavation as a conservation measure. Most of the technical as well as institutional aspects of manufacture and manuscripture still await proper investigation, but the simple point that bears stressing here is that with clay tablets the preparation of the writing material goes hand in hand with the writing itself. Once the clay is dry, further writing is not altogether impossible (some of Assurbanipal’s acquisitions have their new ownership scratched or ink-written in imitation cuneiform), but the primary text is final; the making of a clay tablet text is a from-start-to-finish business. Text has to fit tablet, but tablet may be made with textual fit in view. The mutual accommodation of tablet and text is another subject in need of more investigation than it has so far received.
Tablet size and shape vary considerably, as also do size and spacing of the writing, which can be quite astonishingly small and compact. The one consistency, for ‘‘epic’’ texts at least, is that the tablets are approximate rectangles, the writing linear. They are sometimes exaggeratedly tall and narrow, sometimes more nearly square, sometimes broader than tall, not often exceeding 30 cm on the longer axis; thickness and curvatures are likewise variable. The writing is on both sides of the tablet (a point of affinity with the codex as opposed to the scroll), and it habitually starts at the very top and occupies the full width of the tablet, any empty area coming only at the end of the tablet’s text. The text may be disposed in a series of columns, two or three or more per side, close-set; the wall-to-wall furrows present a very different appearance from the parade of columns through the length of a papyrus scroll, black blocks of text set in lavish white surroundings. Columns are routinely taller than they are wide, often about twice as tall or more, but the relative as well as the absolute dimensions show considerable variation. Both single-column (per side) and multi-column tablets were in coexistent use in many different eras and places, with correspondingly different shapes. In some cases it is possible to discern formats which may have been more or less standard: tall narrow tablets for Enuma Elish, large broad multi-column ones for Gilgamesh (whether or not in its first-millennium ‘‘Standard Babylonian” form, Sha naqba imuru), but such constancies may not extend very far. The tendency overall is for tablets to get bigger, texts to get longer, columns of writing to get broader, but such sweeping generalizations are subject to many qualifications. An Old Babylonian text of the Atrahasis epic found at Sippar, dating from ca. 1700 BCE, was written on large tablets (25 cm high x 19.5 cm broad) with four columns per side; a text of the ‘‘Standard Babylonian’’ version of the same epic, also from Sippar but produced about a thousand years later, occupied narrow tablets (17 x 7 cm) with only one column per side.
Such variations aside, the medium is very stable. The texts copied for the palace collections at Nineveh under Ashurbanipal show the ultimate state of the art, palpably inviting admiration while at the same time being wholly conventional: large, very well made tablets with straight edges, squared corners, and smooth plane surfaces, not square but variously oblong, the writing calligraphically executed in ruled columns with very narrow gutters. The unprodigal use of space, which contrasts with the spaciousness of deluxe Greek and Latin manuscripts and is only accentuated by the occasional intrastichic gaps, is here an item of refined showmanship. As thoroughly aestheticized products they do their job, or much of it, without needing to be read. The inscribed cylinders and other such artifacts in the palace show the same or analogous features, contrived perfection rendered in the economy of fit between the writing and the recipient object.
However big the tablet and small the writing, few tablets carry more than three or four hundred lines. And unlike a scroll, which can always have more added to it if it turns out not to be long enough for its assigned text, a tablet is not extensible. It has been suggested that the commonest size of tablet in the Old Babylonian period, ca. 13 x ca. 9 cm, carrying about 30-5 lines on each side, is responsible for articulation of Sumerian poetic compositions in multiples of around 60-70 lines (Vanstiphout 1995: 2184). There is room for skepticism here, for Sumerian poems may not show sufficient consistency in either length or structuring, but it is undeniable that there is interdependence between tablet and text, and there clearly has to be some form of negotiation between the size and shape of the one and the size and shape of the other. It could be that the text is controlled by the medium not only in the form of its presentation but also in a more constitutive sense. At all events, the question of their interaction is much more significant than is the case with scrolls.
Whatever the inhibitions, a text can be split up and divided between two tablets, or more than two. Once text and tablet cease to be physically coextensive, limits on length are removed. At the same time, once a text is split up, its integrity is lost, or at least jeopardized; each tablet is a literally independent entity. Methods of repairing the loss of physical unity were in fact much the same as with modern multi-volume works: the constituent tablets could be of the same size, and match in appearance; there could be labels, on pieces of clay attached by string; and a set or group of tablets could be kept in a container, itself labelled, or in the same ‘‘pigeonhole’’ (this sort of storage system is preserved at Khorsabad and Sippar, the walls lined with tiers of niches or minicubicles), or simply adjacent on the same shelf. Internally, correct sequence is facilitated by catchlines, i. e. by including the first line of the next tablet at the end of the current tablet’s text, ruled off fTom it or not. Most significantly of all, there was what may be termed metatext, in the form of the colophon, a subscription appended at the end of the main text. Its first and most indispensable item is the tablet’s self-identification: ‘‘tab. 3, Enuma elish’’ (the poem’s first words serving in effect as title). By these means a text’s sundered members could be united. A reader (scribe/librarian/priest/scholar/functionary) would know what tablet came next in the series (as well as what came first), whether or not he could find it.
The organization of multi-column multi-tablet texts must have required quite careful planning. The Old Babylonian Atrahasis mentioned above, for example, has 1,245 lines, distributed over three tablets each with four columns per side: the arrangement had to be worked out in advance, so as to ensure that each tablet’s text would end in its eighth column, with enough room left for the colophon. This particular set of tablets, like many others, registers the line-count of its text, tablet by tablet: every tenth line is marked (line-initial wedge for ‘‘10’’), and the colophon of each tablet, after identifying its membership of the set, records its line-count (416 lines for tab. 1,439 for tab. 2,390 for tab. 3 and then ‘‘total 1,245, 3 tablets’’). The primary function of the line-count was no doubt as a control on textual integrity, but it may have incidentally assisted the distribution into columns and tablets. Things did not always go well: sometimes a tablet turned out to be not quite big enough for the text it was assigned to hold, forcing the scribe to compress the layout and use the tablet edges in order to get it all in. Worse miscalculation would be ruinous. Such exigencies do not apply to scrolls, but return with the codex.
The columns of text on either side of the tablet are arranged in a way that may seem odd to us, and would certainly have seemed odd to Greeks and Romans. On the front of the tablet (the ‘‘obverse’’) everything proceeds in familiar fashion: the direction of writing is from left to right (a third-millennium innovation), the sequence of columns likewise. But on the other side (the ‘‘reverse’’), while the direction of writing is still fTom left to right, the sequence of columns is from right to left; and the text is upside-down relative to the text on the front. The upper lefthand corner of the obverse, the beginning, will be the lower lefthand corner of the reverse, the end. It all seems less odd once the implied procedure is recaptured. On arriving at the end of the last column of the front, the scribe - or the reader - turns the tablet over, from top to bottom, and as it were continues the column; the remaining columns then proceed leftwards. It is a standing convention of the medium.
A composition’s poetic form is normally given recognition by the layout of the text, indeed is virtually constituted by it. In the main Mesopotamian tradition, Sumerian and Akkadian, the written text consists of stichoi, lines of verse, each of which is at the same time a sentence or something like it, sense-units and verse-units being coextensive. In the absence of a self-defining metrical form such as the dactylic hexameter, analyses have trouble avoiding circularity, but each line of text is what is conventionally termed a verse. Lines are right-justified as well as left-justified (that is to say, they not only begin but also end in vertical alignment), intensifying a visual effect further enhanced by the columns’ being set very close to one another and often marked off by vertical rulings to left and right. This contrasts with the practice in Greek and Latin poetic manuscripts, which conspicuously make no attempt to regularize the line-length of metrically equal lines. Within the line, too, the practice differs. Whereas in Greek and Latin manuscripts the verse is written uniformly letter by letter from beginning to end, in Mesopotamian there is intrastichic articulation. The verses often fall into two halves (hemistichs), and the written tradition often indicates the medial segmentation (caesura); in some manuscripts the second halves are set in vertical alignment, just as the line-beginnings are; there is little inhibition against resultant gaps within the line. In short, there is correlation between the formal layout of the lines on the one hand and their poetic structure in terms of syntax and sense on the other, both at the level of the verse (with the privilege of a line to itself) and also within the verse (the medial articulation). This is how the song is scribally actualized. Equivalent oral rendition can be imagined (substituting temporal actualization for spatial, vocal for visual). The written poem is self-standing.
The Ugaritic ‘‘epics’’ (see Chapter 16, by Wyatt), curiously, do not observe this tradition of stichic layout. Though internally they appear to show comparable verse-forms, and were written on multi-column tablets, they are written simply as continuous text, not in lines that correspond to the verses, or only intermittently so. They are not acknowledged as poems at all. The same is true of Hittite texts (see Chapter 17, by Beckman). There may be special factors at work in each case. The Ugaritic texts are single-copy specimens, most if not all of them the product of a single scribe (the colophons identify him), and there is no way of judging their typicality. The Hittite texts, for their part, are versions of Hurrian originals, and their status as verse seems open to question. Texts of Akkadian epic at Hattusas follow the same system of layout as elsewhere.
Justified verse layout calls for lines far from equal in the number and form of their constituent characters to occupy the same extent of horizontal space. In practice it is not feasible, and the system not rigorously sustained. Often a long line cannot be accommodated within the alloted space and spills over, or when columns are narrow verses may be divided over two lines; conversely, verses are sometimes doubled up (especially when the scribe realizes he will otherwise run out of space on the tablet), or inconsistently lineated. But scribes do have means of achieving the requisite equalization, by elongating the component signs of the shorter verses, by leaving gaps between their character-groups; such leveling can be done unobtrusively, masking the inequalities, but preferred practice is to leave pronounced gaps. In ostentatiously calligraphic manuscripts it appears to be something of an affectation, these intermittent blanks on an otherwise fully inscribed surface standing out like bursts of sudden silence.
The verse is normally the largest poetic unit accorded textual recognition, but horizontal rulings sometimes mark off larger structures. They sometimes mark speech termini (as paragraphi sometimes do in Greek manuscripts), sometimes effect other, less organic demarcations; sometimes they occur at points in the text where other manuscripts pass from one tablet to another, and in such cases that is perhaps just what they are indicating.
Soft clay is not the only substance that will take and hold an impression, and so be suitable for cuneiform. There is also wax. The waxed writing-board - sometimes referred to as a stylus tablet, to distinguish it from the hard-surfaced pen-and-ink tablet - in fact outlasted the clay tablet, for it successfully made the transition from the cuneiform world of the Near East to the non-cuneiform world of Greece and Rome. It is true that non-cuneiform scripts can be written with a stylus in clay as well as in wax, and clay tablets were routinely used for Greek in the Mycenaean age, as the thousands of Linear B tablets testify. But once Greek had its alphabet, there was little further place for clay, except in the very different form of the ostrakon, broken pottery. The wax tablet, however, retained its utility. In the Greco-Roman world it was not used as a medium of literature in other than draft or exercise form; its great advantage is that it can easily be used over and over again. But in the Near East it was employed not only for day-to-day purposes but also for library texts. Actual survivals are next to none, but colophons of surviving clay-tablet copies of texts sometimes attest writing-board exemplars, and inventories of accessions to the royal libraries at Nineveh under Assurbanipal catalogue writing-boards along with clay tablets.
The board was ordinarily of wood, one or both sides shallowly recessed for the writing area, which was filled with a thin layer of wax. The wax itself was actually a mixture of beeswax and orpiment (trisulphide of arsenic), a mixture which is more workable and gives a more plastic writing surface than wax alone and has a more attractive yellow color; in the Greco-Roman sphere pigments of other colors were used. The orpiment and in some regions the wood would have to be imported, but widespread use is attested from the second millennium on, in a variety of woods.
Boards can be used singly. But it is the simple fact that they can be joined together that in terms of mediumistic potential crucially distinguishes tablets of wood (whether wax-surface or hard-surface) from tablets of clay. The diptych (to use the Greek term) is a pair of boards attached in such a way that it can be opened and closed. The writing was on the two facing sides, exposed to view when the diptych was opened, concealed and protected when it was folded shut. The ‘‘folded board’’ with its ‘‘lethal signs’’ that Bellerophon took with him from Greece to Lycia in Iliad 6, the only mention of writing in Homer, shows at least indirect acquaintance with the medium. What must be an exceptionally fine example was recovered from the fourteenth-century Uluburun shipwreck off the Lycian coast. Less than 10 cm in height, with a correspondingly small writing area of ca. 7.5 x ca.4.5 cm, this pair of boxwood boards had ivory hinges (three hinges, at least two of them ivory), equipped with hidden wooden dowels and attached to the grooved edges of the boards by hidden wooden nails; rotation through 180° enabled the boards to lie flat when open, a loop-and-hook mechanism fastened them closed. All in all it is a fairly elaborate piece of carpentry. Normally, one may imagine, the construction would be simpler, the boards linked by cords passing through bored holes, as later at Rome; but it could be that the diptych had a status that called for less crude styles of manufacture.
The number of boards is not limited to two. Several of the writing-board accessions itemized in the Nineveh catalogues consisted of more than four. The distinction between ‘‘diptych’’ and ‘‘polyptych’’ used in modern discussions does not correspond exactly to Akkadian classification, where the (waxed) writing-board medium is generally designated le’u, sometimes specified as wood (the kind of wood sometimes also specified; other materials are ivory and lapis lazuli), and an individual board is referred to as a daltu or ‘‘door’’ (the word comes into Greek as deltos, carrying the writing-board signification); a writing-board set consists of a given number of‘‘doors.’’ In the context of the codex - to shift again to later terminology - these are ‘‘leaves’’ (Lat. folia), a term more apt for parchment than for unpliable wood. The assumption might be that the multi-board sets were constructed in the familiar form whereby the boards are linked all at one side, as in a codex. But a remarkable survival suggests otherwise. A late eighth-century polyptych discovered at Kalhu, made not of wood but of ivory and consisting of no fewer than sixteen boards, was put together by hinging the boards one to the other concertina-style, ‘‘like the separate leaves of a Japanese screen’’ (Howard 1955). Admittedly this was no ordinary book; the level of craftsmanship is phenomenal, especially in the sophistication of its hinging, and no ordinary book was ever made of ivory. In fact, as its cover-title proclaims, it was made for the king of the world, Sargon II, destined (or rather intended, for it never got there, or if it did was promptly purloined) for his new palace at Dur-Sarrukin. Even so, it would be surprising if there is anything new about its basic make-up, and it is best taken as a maximally luxurious rendering of a conventional form. It seems quite possible, then, that this concertina method of construction of multi-board sets was standard. The boards would both unfold and fold very neatly, and it is certainly no less ‘‘natural’’ a way of extending a diptych than the alternative codex method, which is without Near Eastern attestation. Indeed, if the first stage of development is thought of in terms of joining two diptychs together, the concertina arrangement is the only way of having the written sides directly succeed one another. This kind of construction is known in other cultures (China, most notably), and it is interesting to find it independently resorted to in the case of a set of thin folded wooden leaves used by Roman military personnel in Britain around the end of the first century ce (Bowman and Thomas 1983: 40, 83-4). This concertina form of book has no established name; it could be called a ‘‘ptych.’’
The Sargon ptych is a book that can be judged by its cover, which grandly displays a centered and boxed four-line inscription incised in the ivory, identifying the thing: ‘‘Palace of Sargon, king of the world, / king of Assyria. The text series (beginning) Enuma Anu Enlil / he had written on an ivory writing-tablet and / deposited it in his palace at Dur-Sarrukin’’ (trans. Wiseman 1955). The very existence of covers, and potentially of cover-titles, is a feature of the medium inherently alien to the clay tablet. A clay tablet has only two sides, which is to say it consists only of front and back (and edges), and the text always begins at the very top. A ptych, like a codex, has not only front and back, which are now exterior sides, but interior sides too, and the exterior ones can serve as covers to protect the text written on the interior. Wooden diptychs, such as the Uluburun specimen mentioned above, had only the inner sides recessed to receive the wax and the writing, and Sargon’s polyptych follows suit. How usual it was to utilize the otherwise vacant front cover for identification is impossible to say, but the medium certainly lends itself to that. This stands in sharp contrast with clay-tablet practice, where all such metatextual data are deferred to the end, in the colophon.
The layout of the text on writing-boards may be assumed to have followed the model of the clay tablet, but the Sargon ptych is the only surviving specimen on which it can be observed. This particular ptych, perhaps unusually large (though wooden boards found in the same cache were somewhat larger), was about twice as tall as it was wide (ca. 34 X ca. 15.5 cm), a common shape for clay tablets, and the impression of height was enhanced by the text’s being laid out in two long narrow columns per board-side, about 125 lines to the column, the writing small. The whole book, if it did indeed comprise the sixteen boards as reconstructed, would have accommodated around 7,500 lines. That is more than epic length. The text, however, is a well-known astronomical omen series; copies of such texts, embodying the forms of Babylonian scientific prognostication, hugely outnumber copies of epics. The fact remains that in the ptych we meet a high-prestige medium with the capacity to contain whole epics.
It was on clay that Enmerkar had invented writing, and one might have imagined that the venerable clay tablet was always the more authoritative medium, immutable and ideologically pre-eminent. Whether such a view can survive the challenge posed by the Sargon ptych seems to be an open question.
World History
