discoveries came thick and fast.29 Our new understanding shows that there were in Mesopotamia several forms of 'proto-writing' in use before writing proper. Of these, stone cylinder seals were both more permanent and at the same time more flexible versions of the clay 'envelopes' examined by Schmandt-Besserat. The seal itself took the form of a hollow cylinder, on which was inscribed a set of engravings. The cylindrical seal would be rolled over wet clay, which therefore reproduced the engraved inscription as a reversed, embossed image.30 The clay seals were used everywhere: they could be moulded over the knot of a rope tied around a bundle; or over the rope fastening of a door. The idea was that the seal should bear a clear mark, identifying its owner.31 Like the clay envelopes studied by Schmandt-Besserat, seals were instruments of economic control, guaranteeing the supervision of proceedings, or confirming that a transaction had taken place. In practice, the Sumerians produced some very imaginative devices with which to identify owners: worshipping at a temple, processions of boats, prisoners before a ruler, the feeding of animals. They were, in effect, pictographical signatures.32 Later, a new type of seal emerged, produced by cutting machines. This clearly suggests that trade was increasing and that the need for identifying marks was likewise growing. So much for proto-writing. But cuneiform actually developed out of the archaic Uruk pictographic system, which took over many of the signs used with the earlier tokens, such as the sign for sheep, and wavy lines for water. The birth of writing proper is clearly shown by the use the first scribes made of the so-called 'bevel-rimmed' bowls of Uruk. These were cheap, coarse and very porous. They could not have been made to hold water and yet they were so common that, at some sites, they made up three-quarters of the pottery found. The fact that they were so porous-suitable only for containing solid matter-and were all the same size, provides a key to their use. Texts that have been deciphered tell us that the workers of Uruk, at least the workers on the large temple projects, were paid in kind-i.e., with a daily ration of food. Since the bulk of the workers' rations would have been grain, it stands to reason that these were the 'standard' bowls by which the workers were paid.33 Shown in Figure 4, opposite, is the very ancient sign for 'eat'. This quite clearly shows a head, with an open mouth, receiving food from one of these 'bevel-rimmed bowls'. It was, in other words, a picture, or pictograph. Many other words began as pictographs, too (see Figure 5, opposite). Figure 4: A bevel-rimmed bowl and the early sign for 'to eat' (left); as it begins to be represented in early cuneiform (right) 34 [Source: Hans J. Nissen, The Early History of the Ancient Near East: 9000-2000 BC , translated by Elizabeth Lutzeier with Kenneth J. Northcott. 1988 by the University of Chicago] This was only the beginning. Just as cylinder seals became simpler and easier to mass-produce-to cope with busy life-so too did writing evolve. Writing on moist clay made it awkward to draw these images clearly and quickly (a problem which the Egyptians never had, with their smooth, dry surfaces, which is why they stuck with hieroglyphics), and so signs, words, became more abstract, fewer, aligned much more in the same direction, all developments that enabled the speed of writing to be increased. Figure 6 on page 82 shows how a few words changed in appearance, over a millennium and more, from the earliest days in Uruk, to the height of Ur's power, that is, between c . 3800-3200 and c . 2800-2100 BC. We still don't know why the images were turned through ninety degrees, but this would surely have made the images less legible and that in turn may have provoked a more simple way of writing. Circular and curved marks were always more difficult to produce in wet clay and this is why cuneiform emerged as a system of simple strokes and wedges. The repertoire of signs was reduced and homogenised by the first third of the third millennium.
Figure 5: Early pictographs: (a) a group of reeds; (b) an ear of corn; (c) a fish; (d) a goat; (e) a bird; (f) a human head; (g) a form of pot; (h) a palm tree; (i) a ziggurat 35 [Source: H. W. F. Saggs, Civilisation Before Greece and Rome , London: B. T. Batsford, 1989, page 62] Figure 6: The development of pictographs into Babylonian cuneiform script 36 [Source: Hans J. Nissen, The Early History of the Ancient Near East: 9000-2000 BC , translated by Elizabeth Lutzeier with Kenneth J. Northcott. 1988 by the University of Chicago] In these early phases, the uses of writing were limited and, because of its basis in trade, consisted just as much of numbers as of words. Among the signs, for example, there was one which had a D-shape: there was a straight edge which was deep-cut and a round end which was much shallower, reducing to nothing. What gave the game away was that these Ds were grouped into clusters, ranging from one to nine. Here then was the making of a decimal system. In some cases, a circular punchhole, formed by means of a cylindrical reed pressed into the clay, was associated with the Ds. 'It is a reasonable assumption that these "round holes" represent tens.'37 It was common for the early tablets to have a list of things on one side, and the total on the other.38 This helped decipherment. A system of signs was one thing. But, as we have seen in examples from elsewhere, such a system does not fully amount to writing as we know it. For that, three other developments were necessary: personal names, grammar, and an alphabet. Personal identification was a problem and a necessity from the moment that economic organisation went beyond the extended family, where everyone knew each other and property was owned communally. Certain names would have been easy, 'Lionheart' say.39 But how would one render an abstract name, such as 'Loved-by-God'? Pictographs would have been developed, much as the heart shape, , has come to mean 'love' in our time. In this way, multiple meanings overlapped: the sun, , for example, might mean 'day', 'bright', or 'white', while a star, , might mean 'god' or 'sky', depending on context. The 'doctrine of the name' was important in Babylon, where thought worked mainly by analogy, rather than by inductive or deductive processes as we use in the modern world.40 For both the Babylonians and the Egyptians the name of an object or a person blended in with its essential nature.41 Therefore, a 'good' name would produce a 'good' person. For the same reason, people were named after the gods and that was also the case with streets ('May the enemy never tread it') and canals and city walls and gates ('Bel hath built it, Bel hath shown it favour'). To cap it all, the practice evolved to adopt a certain tone when uttering proper names. This was especially true when speaking gods' names and it is still true today, to a certain extent, when people use a different tone of voice when praying out loud.42 To begin with, there was no grammar. Words-nouns mainly, but a few verbs-could be placed next to one another in a random fashion. One reason for this was that at Uruk the writing, or proto-writing, was not read, as we would understand reading. It was an artificial memory system that could be understood by people who spoke different languages. Writing and reading as we know it appears to have been developed at Shuruppak in southern Mesopotamia, and the language was Sumerian. No one knows who the Sumerians were, or where they originated, and it is possible that their writing was carried out in an 'official' language, like Sanskrit and Latin many thousands of years later, its use confined only to the learned.43 This next stage in the development of writing occurred when one sound, corresponding to a known object, was generalised to conform to that sound in other words or contexts. An English example might be a drawing of a striped insect to mean a 'bee'. Then it would be adapted, to be used in such words as 'be-lieve'. This happened,