= kho-pi-ru = to be, where the sounds and not the meaning of the several components are alone attended to[115].
The Alphabet.
By analogous processes was formed a true alphabet, in which, however, each of the phonetic elements was represented at first by several different characters derived from several different words having the same initial syllable. Here was, therefore, an embarras de richesses, which could be got rid of only by a judicious process of elimination, that is, by discarding all like-sounding symbols but one for the same sound. When this final process of reduction was completed by the scribes, in other words, when all the phonetic signs were rejected except 23, i.e. one for each of the 23 phonetic elements, the Phoenician alphabet as we now have it was completed. Such may be taken as the real origin of this system, whether the scribes in question were Babylonians, Egyptians, Minaeans, or Europeans, that is, whether the Phoenician alphabet had a cuneiform, a hieroglyphic, a South Arabian, a Cretan (Aegean), Ligurian or Iberian origin, for all these and perhaps other peoples have been credited with the invention. The time is not yet ripe for deciding between these rival claimants[116].
The Persian and other Cuneiform Scripts.
But whatever be the source of the Phoenician, that of the Persian system current under the Achaemenides is clear enough. It is a true alphabet of 37 characters, derived by some selective process directly from the Babylonian cuneiforms, without any attempt at a modification of their shapes. Hence although simple compared with its prototype, it is clumsy enough compared with the Phoenician script, several of the letters requiring groups of as many as four or even five "wedges" for their expression. None of the other cuneiform systems also derived from the Sumerian (the Assyrian, Elamite, Vannic, Medic) appear to have reached the pure alphabetic state, all being still encumbered with numerous complex syllabic characters. The subjoined table, for which I have to thank T. G. Pinches, will help to show the genesis of the cuneiform combinations from the earliest known pictographs. These pictographs themselves are already reduced to the merest outlines of the original pictorial representations. But no earlier forms, showing the gradual transition from the primitive picture writing to the degraded pictographs here given, have yet come to light[117].
The Mas-d'Azil Markings.
Here it may be asked, What is to be thought of the already-mentioned pebble-markings from the Mas-d'Azil Cave at the close of the Old Stone Age? If they are truly phonetic, then we must suppose that palaeolithic man not only invented an alphabetic writing system, but did this right off by intuition, as it were, without any previous knowledge of letters. At least no one will suggest that the Dordogne cave-dwellers were already in possession of pictographic or other crude systems, from which the Mas-d'Azil "script" might have been slowly evolved. Yet E. Piette, who groups these pebbles, painted with peroxide of iron, in the four categories of numerals, symbols, pictographs, and alphabetical characters, states, in reference to these last, that 13 out of 23 Phoenician characters were equally Azilian graphic signs. He even suggests that there may be an approach to an inscription in one group, where, however, the mark indicating a stop implies a script running Semitic-fashion from right to left, whereas the letters themselves seem to face the other way[118]. G. G. MacCurdy[119], who accepts the evidence for the existence of writing in Azilian, if not in Magdalenian times, notes the close similarity between palaeolithic signs and Phoenician, ancient Greek and Cypriote letters. But J. Déchelette[120], reviewing (pp. 234, 236) the arguments against Piette's claims, points out in conclusion (p. 320) the impossibility of admitting that the population of Gaul could suddenly lose so beneficial a discovery as that of writing. Yet thousands of years elapse before the earliest appearance of epigraphic monuments.
Evolution of the Sumerian Cuneiforms.