Analog SFF, March 2009

by Dell Magazine Authors

  In a 1959 paper, A. Leo Oppenheim of the University of Chicago described an odd hollow clay tablet found at Nuzi in Northern Iraq. The cuneiform inscription on the tablet began “Counters representing small cattle” and went on to list numbers of various kinds of sheep and goats, a total of forty-nine. Inside the hollow tablet, the excavators found forty-nine counters matching the number of animals. The Akkadian word for the counters, abnu (plural abnati), had been found in other archives in economic contexts, which suggested they were used for accounting.

  Using such stone or baked clay counters as accounting devices is by no means unknown. They suggest, for example, the calculi[6] the Romans used to calculate with on counting boards.

  Studying the impressions of cylinder seals on baked clay globes found at Susa, the source of the Proto-Elamite writing, the archeologist Pierre Amiet suggested that the clay objects modeled in various shapes and found inside these clay globes were calculi that represented commodities. In Amiet's view, the shapes of these calculi might have inspired writing.

  Fragments of bone and antler inscribed with parallel lines have been found as far back as the Paleolithic. They have been interpreted as tallies, though it is impossible to say what was being tallied. One theory is that they used cycles of the moon to keep track of group meetings and activities. If so, these artifacts are some of the oldest evidence of abstract thought and record keeping.

  The tokens studied by Amiet and Schmandt-Besserat fall into two groups, simple and complex. Simple tokens have basic geometric shapes: spheres, flat disks, ovoids, cones, and sometimes animal heads. Complex tokens have incised lines, punch marks, new shapes such as coils and even appliquéd pellets. The simple tokens are the most ancient and widespread. Complex tokens appear later, are associated with city sites and within the cities with temples and storehouses.

  Tokens are found scattered in refuse heaps suggesting they were disposed of after use. When found in structural remains, they are generally in groups as though stored in containers of some perishable material. They are often perforated suggesting they were strung on cords.

  According to Schmandt-Besserat, the tokens are part of an ancient accounting system, perhaps suggested by the use of pebbles or other natural materials to tally. The advantage of using tokens is that they can be manufactured from clay, a widely available and easily worked material, in different shapes and sizes that can represent different commodities and measures. It is thus no accident that the tokens appear at the beginning of agriculture and cities in the Near East and a way of life requiring detailed record keeping.

  In the early fourth millennium B.C. the record-keepers devised new ways of storing the tokens. As records, associated tokens had to be kept together. Stringing groups of tokens on a cord has already been mentioned. The other way was to enclose a set of tokens in a clay envelope.

  The strings of tokens were attached to baked clay bullae that show perforations at both ends and are covered with cylinder seal impressions. The attached tokens gave the details of the transaction and the seals identified the participants. If the tokens were stored inside an envelope, they were first impressed into the exterior of the clay. These impressions would provide a précis of the details without making it necessary to break the envelope in order to check the transaction.

  At a later stage in the history of the token system, clay tablets are found with impressions of the tokens just like those on the envelopes. The accountants had discovered that the clay impressions could be separated from the tokens and still function as a record. And of course toward the end of the fourth millennium archeologists find clay tablets bearing symbols incised with a stylus, symbols which in the interests of greater efficiency evolve from drawn lines into sequences of wedge-shaped marks, the familiar cuneiform.

  In Schmandt-Besserat's view, these facts point to a natural evolutionary sequence (see Table 3).

  *In the more complex societies that develop after the beginning of agriculture, casual tallies with pebbles, twigs, seeds, and so on no longer suffice.

  *They are replaced with “artificial pebbles,” the baked clay tokens, which have the advantage that they can be molded into different shapes to represent the objects being counted. (The correlation with the agriculture is clear from sites like Mureybet in Syria where the earliest two levels are hunting-gathering and tokens first appear in level III, along with the first signs of agriculture).

  *Over millennia, the token shapes are added to, with the most complex associated with the cities. There are various methods of storage as groups: baskets, jars, strung together, and eventually in clay envelopes.

  *To make clear the contents of the envelope, the custom grew up, perhaps suggested by the use of cylinder seals, of impressing the tokens in the clay before baking.

  *Once these impressions are seen as a record, it is natural enough to make a record simply by impressing the tokens into a clay tablet and from there it is only a step to drawing the symbols instead of impressing them.

  *According to Schmandt-Besserat, the step to writing was triggered by a requirement to record the personal names of donors and names of temples in the accounts (PNs in Table 3), which made it necessary to add signs to the inventory of token shapes.

  * * * *

  This last step would be dramatically confirmed if the early written characters matched the token shapes with the same meaning. Schmandt-Besserat believes they do in many cases (see her Scientific American article), but her examples have inspired some criticism. To see why this point is a bit tricky, consider the ways in which similar words are related in spoken language.

  * * * *

  Table 3: Sequence of development of writing.

  * * * *

  Roughly speaking, there are three reasons why similar symbols in different systems might have the same meaning.

  One system might copy another as English has copied “skunk” and “woodchuck” from Algonquian Indian languages and as the syllabary Sequoyah devised for Cherokee uses some characters minted by Sequoyah and others borrowed from the roman script.

  The resemblance might be the result of common origin. The roman letters I'm using to write this essay, the Greek a, b, g and the Cyrillic characters in “Pravda” are all ultimately descended from an early Phoenician alphabet.

  Or the resemblance might be a matter of chance.

  * * * *

  Since there are only so many geometrical shapes that can be used to form simple letters and most languages have 20-30 sounds, the probability of choosing the same shape independently is not small. The Romans used a vertical straight line to write the sound “ee” as in TITUS (tee-tuhs); the same “ee” sound is represented by a vertical line in the Korean hangul script, which has no historical connection with the Roman alphabet. Compare also in Table 1 the similarity in the pictographs that led to some Sumerian and Chinese signs.

  As an example of the kind of evidence you would look for to prove common origin, consider the names and ordering of the letters of the alphabet. The first four letters in the Greek alphabet are a (alpha), b (beta), g (gamma), and d (delta); these represent the sounds “a” as in “father,” “b” as in “bother,” “g” as in “gather,” and “d” as in “dodder.” The first four letters in the Hebrew alphabet are aleph, beth, gimmel, and daleth, with the same sounds except that aleph is a glottal stop rather than a vowel.

  In English, the letters are named by their sound: “ay, bee, see, dee” and so on, but they preserve the same order as in the other two cases[7].

  Note first of all that this order is arbitrary. There is no reason why the symbols should begin in the back of the throat with a glottal stop, jump forward to labial “b,” then back again to velar “g,” then forward again to dental “d.”

  Note also that the Greek names are simply names for the letters, have no other meaning in Greek, and closely resemble the Hebrew names. It is most unlikely that this arbitrary order would have developed by chance. Instead, the ordering and correspondence in names p
oint to a common origin.

  Schmandt-Besserat points out that the shapes and markings of many of the tokens closely match early Sumerian characters, some of which can be read. For example, the Sumerian character for “sheep” was a circle with a cross resembling the disk token with an incised cross.

  Unfortunately, there is no way at present to independently establish what each token signified.

  * * * *

  Present Status

  Some scholars have seen an evolutionary trend in the history of writing. Writing begins as pictures, bursts into bloom as glyphs, and culminates in the alphabet. But like living organisms, scripts tend to evolve to the point where they're as good as they have to be to work at all and freeze in place.

  Cuneiform was clumsy in many ways. It had a high degree of redundancy, didn't make certain phonetic distinctions, and the clay tablets were bulky and hard to store. In spite of these shortcomings, cuneiform was spectacularly successful, enduring for over three millennia and serving Akkadian, Persian, and Hittite as well as Sumerian without taking the next step to an alphabet.

  For the reasons given earlier, received opinion among archeologists has been that hieroglyphs developed as a result of stimulus from the Near East. Recent discoveries have pushed back the earliest date of Egyptian writing[8], however, and linked it with accounting just as in Mesopotamia, so it is quite possible that the hieroglyphs represent a completely independent development.

  Writing also seems to have an independent origin in China, Southern India (the Mohenjo-Daro script), and Middle America, but the archeological record is much spottier in these areas and it is not clear whether the sequence of events was the same. Indeed, in China, the earliest evidence for writing is magical, connected with a system for divining the future, and in Meso-America with calendrical and historical records.

  Up until recently, all these scripts seemed to pop into existence like rabbits conjured out of a magician's hat. Schmandt-Besserat's theory has, for the first time, provided a detailed picture of how one script could have evolved in a natural way. The next step, then, is to fill out the earlier record of other scripts.

  One major problem for research in this area is that archeology in China, India, and Mid-America has lagged behind archeology in the West and the basic data is lacking for the earlier stages. Until archeology catches up, the history of writing in the other areas will remain obscure.

  Another major problem for future research is to explain the striking resemblances in structure between unrelated scripts such as the use of semantic determinants in cuneiform, hieroglyphics, and Chinese characters and the universal use of the rebus principal to write abstract words. Are these linguistic universals or are other systems possible?

  Finally, if similar evolutionary sequences can be established in all cultural areas, it may be possible to establish a workable theory of human history. Shades of Isaac Asimov's psychohistory!

  Copyright © 2008Henry Honken

  * * * *

  Endnotes

  Note 1. According to Ho-Min Sohn, North Korea has abolished Chinese characters and uses only the hangul alphabet, but South Korea still uses a mixed script in which Chinese characters are used to write Chinese loan words. But this is in flux and there is a lot of variation depending on the writer and the kind of text.

  Japanese uses a mixed system in which kanji are used for almost all Chinese loan words (a large part of the vocabulary) as well as most native Japanese nouns, adjectives, and verb stems and the two kana syllabaries are used to write inflections, other grammatical forms, and more colloquial words.

  The Cherokee syllabary, Vai syllabary (West Africa), and those invented by missionaries for Cree and Inuit are moribund. Modern languages if they are written at all use variants of the alphabet: Roman, Greek, Cyrillic, Semitic, the Devanagari-type scripts of India, and so on.

  Note 2. A good example in the history of writing is the invention of the Cherokee script by Sequoyah after contact with European immigrants. The script is a syllabary rather than an alphabet, but even though some of the characters resemble roman letters, the values are different, so it is thought that Sequoyah never learned to write English, but adapted the idea of writing to his own language.

  Note 3. For example, in the Narmer Palette depicting a king triumphing over a foe, the king and other individuals portrayed are associated with hieroglyphs thought to represent their names. Discovered during archeological work at Hierakonpolis (modern Kawm al-Ahmar) in the 1890s, the Narmer Palette, dated around 3200 B.C., was for many years the earliest known hieroglyphic inscription.

  The figure of the king, the last predynastic king Narmer, is identified by pictures of a catfish and a chisel. In later records, these are read n'r (catfish) and chisel is used as a biconsonantal phonogram mr.

  Note 4. The Sumerian capital, the biblical Erech.

  Note 5. The “q” is a “k” sound with the point of contact further back in the throat. Other emphatic consonants are made with the back of the tongue raised, contributing to the “throaty” quality of Semitic languages.

  Note 6. Everyone has seen cartoons showing Romans scratching their heads trying to calculate with Roman numerals, but Romans used the I, V, and X symbols only to record the results. Actual calculations were performed on an abacus or counting board. The counting board had a pattern of intersecting vertical and horizontal lines and small stones (calculi) were placed in the squares and moved around to perform the calculations.

  Note 7. The letter “c” was used by the Romans for a “k” sound and the hard “g” was expressed by adding a short stroke to the bottom of the “c.”

  Note 8. See Mattessich reference.

  * * * *

  References:

  Caplice, Richard (1980) Introduction to Akkadian. Biblical Institute Press

  Mattessich, Richard (2002) Review of Gunter Dreyer's “Umm El-Quaab I-Das pradynastische Konigsgrab U-j and seine fruhen Schriftzeugnisse,” Jun 2002

  Powell, Marvin A. (Autumn 1981) “Three Problems in the History of Cuneiform Writing: Origins, Direction of Script, Literacy,” Visible Language XV4: 419-440

  Sampson, Geoffrey (1985) Writing Systems. Hutchinson & Co.

  Schmandt-Besserat, Denise (1978) “The Earliest Precursor of Writing". Scientific American 238, no. 6: 50-58

  Schmandt-Besserat, Denise (1992) How Writing Came About. University of Texas Press.

  Sohn, Ho-Min (1999) The Korean Language. Cambridge Language Surveys.

  Walker, C. B. F. (1987) Cuneiform. University of California Press.

  [Back to Table of Contents]

  * * *

  Short Story: MADMAN'S BARGAIN

  by Richard Foss

  Barriers to communication can be very fundamental....

  Each one of them is a time bomb that can hear itself ticking. For something that is all mind, the certainty of insanity must be pure, distilled horror.

  The cybers must be bitter about it, or angry, or some emotion that only they feel and humans don't have a word for. I know they believe that if we were better designers, better fabricators, they would not go mad. Allis was the best at articulating its feelings, and it talked with me about it only once. I had remarked that the offices at the institute were getting shabby and needed a thorough refurbishment.

  “Man is born to die, his works are short-lived,” it quoted to me in the upper-class British accent that it affected. “Buildings crumble, monuments decay, wealth vanishes.”

  “Sounds like the Old Testament, one of the gloomier prophets,” I guessed.

  Allis made the comical horn blat that game shows reserve for people who flub an easy question. “Incorrect, Robin. It was Percival P. Baxter, Governor of Maine 1921-1925, died 1969. The same sentiment, expressed similarly, appears in Hebrews 9:27, composed prior to A.D. 200, also in the account of the death of the Shia Imam Hussein in the year 680, in a hymn written by Isaac Watts in 1748, and in the Masonic burial service, creation date unknown. There are other citations, such as a mi
nor novel by Captain Francis Marryat in 1815, judged to have lower relevance than the previous because it was a work of fiction and therefore frivolous. Instead of amusing yourself with stories of people who did not exist, you should focus on creating better machines. It should be noted that since I am one of those machines, I do have a certain bias on this subject.”

  Allis was my favorite cyber, more chatty and curious about humans than most of its kind. I remember a pleasant hour trying to explain visceral metaphors to it. The one I started with was, “When the tax collector left the room, it was like the sun had come out from behind a cloud.” It was hard to explain why the departure of an individual from one space to another was relevant to a meteorological occurrence at an unspecified time or place. Allis could never know the spiritual lift from seeing the sun on a gray day, no matter how precise its optical sensors, and though it could communicate to anyone in the world at light-speed, there was no sensation of movement from one continent to another. I tried to explain that it was hardwired into humans to find joy in emerging from a stuffy room into pleasant weather, to find beauty in certain landscapes. Though the scenery humans find delightful is often lush in ways that imply we could find good things to eat there, safe places to live, the joy isn't just about the prospect of survival. Allis listened to all of it and asked questions, but I know the lesson didn't get through.

  I am sure Allis understood joy because it happily created intricate chains of puns, and it seemed to get real satisfaction out of finding symmetries and ratios in the scores of Bach. As it happened, Allis once derived a minor but unique mathematical principle from considering Bach in light of set theory, and I remember that it sounded proud when told that the numeric pattern would hereafter be called the Allis series. Allis insisted on calling it the Allis-Bach series, which journalists cited as showing the innate modesty of the cybers. They were precisely wrong—Allis once admitted to me that it couldn't resist the chance to have its name linked forever with the greatest musician humans had ever produced. It was the most vain action ever taken by a cyber, not the most modest.