At any rate, it was not until 1970 that a young American scholar, K. Leslie Steiner, who had been given an informal account of this parchment by a friend at the University of Tel Aviv, realized that most of the Hebrew words seemed to be translations of words that appeared in that at-one-time most ubiquitous of ancient texts: the Culhar’ fragment.
2
K. LESLIE STEINER WAS born in Cuba in 1949. Her mother was a black American from Alabama; her father was an Austrian Jew. From 1951 on, Steiner grew up in Ann Arbor, where both her parents taught at the University of Michigan, and where Steiner now holds joint tenure in the German, Comparative Literature, and Mathematics departments.
Steiner’s mathematical work has mostly been done an obscure spin-off of a branch of category theory called ‘naming, listing, and counting theory.’ By the time she was twenty-two, her work had established her as one of America’s three leading experts in the field. This was the work that she was shortly to bring to bear on the problem of this ancient text in such a novel and ingenious way. When she was twenty-four, Steiner published a book called The Edge of Language with Bowling Green University Press—not, as one might imagine from our account so far, a treatise on ancient scripts, but rather a study of linguistic patterns common to comic books, pornography, contemporary poetry, and science fiction,* one of the decade’s more daunting volumes in the field of popular and cross-cultural studies. Steiner’s linguistic/archeological interests, nevertheless, have been a consuming amateur hobby—the tradition, apparently, with so many who have made the greatest contributions to the field, from Heinrich Schliemann himself to Michael Ventris, both of whom were basically brilliant amateurs.
Steiner’s recognition of the scroll as a lexicon meant to facilitate the study of the Culhar’ Text in some long-lost language would be notable enough. But Steiner also went on to establish that the language was not Egyptian, at least not any variety we possess. Eighteen months of followup seemed to suggest, from the appearance of the lost script, that, if anything, it was a variety of writing related to the cuneiform ideograms of the Mesopotamian and Indus Valley regions. Her subsequent efforts to locate exactly which form of cuneiform it might be (during which she herself distinguished three distinct forms among the numerous untranslatable tablets that still exist) will no doubt someday make another fascinating book. Suffice it to say, however, that in 1974, one Yavus Ahmed Bey, a 24-year-old research assistant in the Istanbul Archeological Museum, directed Steiner to a codex of untranslated (and presumably untranslatable) texts on store in the library archives.
The codex, a set of loose parchments and vella, had been purchased in Missolonghi in the late summer of 1824, a city and a year that readers of Romantic poetry will immediately associate with the death of Byron—though from all accounts, the sale of the codex, some four months after the poet’s death in the war-ravaged town, had nothing to do with Byron per se. Indeed the 36-year-old poet, who, by the cruel April of his demise, had become obese, drunken, and drug besotted, has the dubious distinction of more than likely knowing nothing at all of the valuable collection of texts that shared the village with him in a basement storage chest a kilometer and a half up the road. The private collector who bought the codex immediately spirited it away to Ankara.
Shortly after World War I, the codex came to the Istanbul museum, where apparently it remained, all but unexamined by any save the odd research assistant. It took Steiner only an afternoon’s search through the contents of the codex to locate the short, five-page text, clearly in the same script as the parchment unearthed thirty years before by a Bedouin youth. Between the Ancient Hebrew lexicon and what is known of other translations of the Culhar’ fragment, it was comparatively simple to establish that here was, indeed, a parchment copy of still another version of the Culhar’, this time in an unknown cuneiform-style language. But the significant point here was a note, in yet another language, written at the end of this parchment; we must point out again that this codex was purchased in 1824 and all but ignored till 1974. But since the late 1950’s, practically any amateur concerned with ancient scripts would have recognized the script of the appended note: it was the ancient Greek syllabary writing from Crete, deciphered by the young engineer Michael Ventris in 1954, known as Linear-B.
The parchment itself, from the evidence of other markings, most probably dates from the third century A.D., but it is also most probably a copy made from a much older source,** very possibly by someone who did not know the meaning of the letters put down. Indeed, it is the only fragment of Linear-B ever to be found outside of Crete. And it is a language that, as far as we know, no one has known how to read for something in the neighborhood of five to six thousand years. The Linear-B fragment, which was soon translated, reads:
Above these words are written the oldest writing known to wise men by a human hand. It is said that they were written in the language of the country called by our grandparents Transpoté.
Here, in this fragment, we most certainly have the explanation for why the Culhar’ was so widespread during ancient times and the nature of its importance: apparently, over a good deal of Europe and Asia Minor, the Culhar’ Text was once thought to be the origin of writing, or the archetrace.
Where Transpoté might be is a complete mystery still, though from internal evidence one would assume it was on a coast somewhere, of a body of water large enough to have islands more than a day’s sail from land. In Greek, ‘Transpoté’ would seem to be possibly a play on the words ‘across never.’ The Homeric meaning includes the possibility of ‘across when’ or ‘a distant once.’ There is also, of course, a more prosaic reading possible, that reads ‘pote-’ as some sort of apocopation of ‘potamos’ meaning river, so that the translation may simply be ‘across the river.’ Other translations possible are ‘far never’ and ‘far when’—none of which, alas, helps us locate the actual country.
But if the Linear-B fragment is authentic, then it establishes with high probability the neolithic origins of the Culhar’ Text—and probably the language transcribed in the Missolonghi Codex—since Linear-B was in use only in the very early stages of the history of the neolithic palaces at Cnossos, Phaistos, and Malliá.
* Steiner has written numerous personable and insightful reviews of science fiction novels that have appeared in several Midwestern science fiction ‘fanzines,’ many of whose readers are probably unaware of her scholarly accomplishments.
** Other parchments in the codex, written in the same ink and presumed to come from the same time, are transcriptions of block-letter Greek inscriptions, that sculptural language written on stone in upper case letters without word-breaks, dating from pre-classic times.
3
BUT TO EXPLAIN THE nature of Steiner’s major contribution, we must leave the Culhar’ fragment itself for a page or so and speak about the origins of writing; and about Steiner’s mathematical work.
The currently reigning archeological theory holds that writing as we know it began not as marks made on paper or skins, or even impressions made on soft clay with pointed sticks, but rather as a set of clay tokens in the shapes of spheres, half spheres, cones, tetrahedrons, and—at a later date—doublecones (or biconoids), as well as other shapes, some with holes or lines inscribed on them, some without. For some five thousand years at least (c. 7,000 B.C. to c. 2,000 B.C.) these tokens in various parts of the Middle East formed a system of account keeping, the various tokens representing animals, foods, jars: and the numbers of them corresponding to given amounts of these goods. The tokens have been found in numerous archeological sites from numerous periods. Until recently archeologists tended to assume they were beads, gaming pieces, children’s toys, or even religious objects. The consistency in the shapes from site to site, however, has only recently been noted. And it was practically at the same time as Steiner was making her discoveries in Istanbul that Denise Schmandt-Besserat realized that a number of the cuneiform signs in the clay tablets associated with Uruk and Nineveh were simply two-dimens
ional representations of these three-dimensional shapes, complete with their added incisions, holes, and decorations.
Thus ‘the violence of the letter’ (a phrase given currency by Jacques Derrida in his book on the metaphor of ‘speech vs. writing’ in Western thought, Of Grammatology [Paris: 1967]) may very well have begun, to use Schmandt-Besserat’s words, with the clay ‘… rolled between the palms of the hand or the lumps pinched between the fingertips … incised and punched.’ Indeed, Derrida’s ‘double writing,’ or ‘writing within writing,’ seems to be intriguingly dramatized by the most recent archeological findings.
In Mesopotamian contractual situations, so runs the theory, these clay tokens were used to make up various bills of lading, with given numbers of tokens standing for corresponding amounts of grain, fabric, or animals. The tokens were then sealed in clay ‘bullae,’ which served as envelopes for transmitting the contracts. The envelopes presumably had to arrive unbroken. In order to facilitate the dealings, so that one would know, as it were, what the contract was about (in the sense of around …?), the tokens were first pressed into the curved outer surface of the still-pliable clay bulla, before they were put inside and the bulla was sealed. Thus the surface of the bulla was inscribed with a list of the tokens it contained. In a legal debate, the bulla could be broken open before judges and the true ‘word’ within revealed.
The writing that we know as writing, in Babylonia at any rate, came about from situations in which such double writing-within-writing was not considered necessary. Curved clay tablets (and the reason for those curves has been hugely wondered at. Storage is the usual explanation. Schmandt-Besserat’s theory: they aped the curve of the bullarum surfaces, from which they were derived) were inscribed with pictures of the impressions formerly made by the tokens. These pictures of the token impressions developed into the more than 1,500 ideograms that comprise the range of cuneiform writing.
Bear in mind the list of tokens impressed on the bulla surface; and we are ready for a brief rundown of Steiner’s most exciting contribution, in many people’s opinion, to the matter. Steiner herself has written in a popular article: ‘Briefly, what I was able to do was to bring my mathematical work in Naming, Listing, and Counting Theory to bear on my archeological hobby. N/L/C theory deals with various kinds of order, the distinctions between them, and also with ways of combining them. In a “naming” (that is, a collection of designated, i.e., named, objects), basically all you can do—assuming that’s the only kind of order you possess—is to be sure that one object is not any of the others. When you have this much order, there are certain things you can do and certain things you can’t do. Now let’s go on and suppose you have a “list” of objects. In a “list,” you not only know each object’s name, but you know its relation to two other objects, the one “above” it in the list and the one “below” it in the list. Again, with this much order, and no more, you can do certain things and cannot do certain others. And in a “count,” you have a collection of objects correlated with what is known as a “proper list.” (Sometimes it’s called a “full list.”) A “count” allows you to specify many, many complicated relationships between one object and the others—all this of course, is detailed in rigorous terms when you work with the theory.’ For the last dozen years or so N/L/C theoreticians have been interested in what used to be called ‘third level order.’ More recently, this level of order has been nicknamed ‘language,’ because it shares a surprising number of properties with language as we know it.
‘Language’ is defined by something called a ‘noncommutative substitution matrix.’ As Steiner explains it, a noncommutative substitution matrix is ‘… a collection of rules that allows unidirectional substitutions of listable subsets of a collection of names. For example, suppose we have the collection of names A, B, C, D, and E. Such a matrix of rules might begin by saying: Wherever we find AB, we can substitute CDE (though it does not necessarily work the other way around). Whenever we find DE, we can substitute ACD. Whenever we find any term following ECB we can substitute AC for that term. And so forth.’ Steiner goes on to explain that these rules will sometimes make complete loops of substitution. Such a loop is called, by N/L/C theoreticians, a ‘discourse.’ ‘When we have enough discursive (i.e. looping) and nondiscursive sets of rules, the whole following a fairly complicated set of criteria, then we have what’s known as a proper noncommutative substitution matrix, or a full grammar, or a “language.” Or, if you will, an example of third level order.’
N/L/C theory got its start as an attempt to generate the rules for each higher level of order by combining the rules for the lower levels in various recursive ways. Its first big problem was the discovery that while it is fairly easy to generate the rules for a ‘language’ by combining the rules for a ‘naming’ and a ‘list,’ it is impossible to generate the rules for a ‘count’ just from a ‘naming’ and a ‘list,’ without generating a proper ‘language’ first—which is why a ‘language,’ and not a ‘count,’ is the third level of order. A ‘count,’ which is what most of mathematics up through calculus is based on in one form or another, is really a degenerate form of language. ‘“Counting,” as it were, presupposes “language,” and not the other way around.’ Not only is most mathematics based on the rules governing the ‘count,’ so is most extant hard computer circuitry. Trying to develop a real language from these ‘count’ rules is rather difficult; whereas if one starts only with the rules governing a ‘naming’ and a ‘list’ to get straight to the more complicated third-level order known as ‘language,’ then the ‘language’ can include its own degenerate form of the ‘count.’
To relate all this to the archeology of ancient languages, we must go back to the fact that we asked you not to forget. Inside the bulla we have a collection of tokens, or a ‘naming.’ On the outside of the bulla, we have the impressions of the tokens, or a ‘list.’
How does this relate to the Culhar’ Text? Soon after Steiner made her discovery in the Istanbul Museum, a bulla was discovered by Pierre Amiet at the great Susa excavation at Ellimite, containing a collection of tokens that, at least in x-ray, may well represent a goodly portion of the words of the ubiquitous Culhar’ fragment; the bulla probably dates, by all consensus, from c. 7,000 B.C. Is this, perhaps, the oldest version of the Culhar’? What basically leaves us unsure is simply that the surface of this bulla is blank. Either it was not a contract (and thus never inscribed); or it was eroded by time and the elements.
What Steiner has done is assume that the Missolonghi Codex is the ‘list’ that should be inscribed on the bulla surface. She then takes her substitution patterns from the numerous versions in other languages. There is a high correlation between the contained tokens and the inscriptions on the parchment discovered in Istanbul.
Using some of the more arcane substitution theory of N/L/C, coupled with what is known of other translations, Steiner has been able to offer a number of highly probable (and in some cases highly imaginative) revisions of existing translations based on the theoretical mechanics of various discursive loopings.
Steiner herself points out that an argument can be made that the tokens inside the Susa bulla may just happen to include many of the words in the Culhar’ simply by chance. And even if it is not chance, says Steiner, ‘… the assignments are highly problematic at a number of points; they may just be dead wrong. Still, the results are intriguing, and the process itself is fun.’
4
WHATEVER OTHER CLAIMS CAN be made for the Culhar’, it is almost certainly among our oldest narrative texts. It clearly predates Homer and most probably Gilgamesh—conceivably by as much as four thousand years.
The classic text in Western society comes with a history of anterior recitation which, after a timeless period, passed from teller to teller, is at last committed to a writing that both privileges it and contaminates it. This is, if only by tradition, both the text of Homer and the text of the Eddas. And we treat the text of Gilgamesh in the same way, though
there is no positive evidence it did not begin as a written composition.
The Culhar’ clearly and almost inarguably begins as a written text—or at least the product of a mind clearly familiar with the reality of writing.
The opening metaphor, of the towers of the sunken buildings inscribing their tale on the undersurface of the sea so that it may be read by passing sailors looking over the rail of their boats, is truly an astonishing moment in the history of Western imagination. One of Steiner’s most interesting emendations, though it is the one least supported by the mathematics, is that the image itself is a metaphor for which might be translated: ‘… the irregular roofing stones of the sunken buildings mold the waves from below into tokens [of the sunken buildings’ existence] so that passing sailors looking over their boat rails can read their presence (and presumably steer clear of them).’ In some forms of the token-writing, Steiner also points out, the token for ‘bulla’ and the token for ‘sea’ are close enough to cause confusion. Steiner suggests this might be another pun.