WRITING (the verbal noun of “to write,” O. Eng. writan, to inscribe), the use of letters, symbols or other conventional characters, for the recording by visible means of significant sounds; more specifically, the art of tracing by hand these symbols on paper or other material, by pen and ink, pencil, stylus or other such means, as opposed to mechanical methods such as printing. The principal features in the development of writing in its primary sense are dealt with in separate articles (see Alphabet, Palaeography, Inscriptions, Book, Manuscript, Shorthand, &c.). Here it is only necessary briefly to refer to the origins of a system which has eventually followed the history of the various languages and has been stereotyped by the progress of typography (q.v.). Very early in the history of mankind three needs become pressing. These are (a) to recall at a particular time something that has to be done; (b) to communicate with some other person who is not present, nor for the moment easily accessible; (c) to assert rights over tools, cattle, &c., by a distinctive mark, or by a similar mark to distinguish one's own production (e.g. a special make of pottery) from that of others. The last-named use, out of which in time develops every kind of trade-mark, is itself a development of the earlier property mark. The right to property must be established before traffic, whether by way of barter or of sale, is possible.

Every one is familiar with devices to achieve the first of these aims; one of the commonest is to tie a knot in a Knot signs. handkerchief. It is obvious that by multiplying the number of knots a number of points equal to the number of knots might in this way be referred to, though it is probable that the untrained memory would fail to recall the meaning attached to more than a very limited number of knots. The simplest application of these knots is in keeping a record of a number of days, as in the story related by Herodotus (iv. 98), to the effect that Darius, on crossing the Ister in his Scythian expedition, left with the Greeks appointed to guard the bridge a thong with a number of knots equal to the number of days that their watch over the bridge was to be continued. One knot was to be undone each day, and if the king had not returned by the time that all the knots were undone, the Greeks were to break down the bridge and go away. A development of this is found in the Peruvian quipus, which consists of a number of thongs or cords hanging from a top-band or cross-bar. In its simplified form, knots are merely tied upon the individual cords. In its more elaborate forms the cords are of different colours, and are knotted together so as to form open loops of various shapes. In the Antiguedades Peruanas,[1] we are told that the knots of the quipus in all probability indicated only numbers originally, but that as time went on the skill of the makers became so great that historical events, laws and edicts could thus be communicated. In every place of any importance there was an official whose business it was to interpret quipus received from a distance, and to make quipus himself. If, however, the quipus which was received came from a distant province, it was not intelligible without an oral explanation. Unfortunately, the art of interpretation of quipus is lost, so that it is impossible to ascertain how far the knots were merely a mnemonic for the messenger, and how far they were intelligible without explanation to a stranger. Similar mnemonics are said to have been used in the remotest antiquity amongst the Chinese, the Tibetans, and other peoples of the Old World.[2]

Similar in character to the quipus is the message-stick, which is still in use amongst the natives of Australia. A branch of a Message-sticks. tree is taken and notches made upon it. These are now cut with a knife; in earlier times they were made with the edge of a mussel shell. The notches are made in the presence of the messenger, who receives his instructions while they are being made. The notches are thus merely aids to memory, and not self-explanatory, though if messages frequently passed between two persons, practice would in time help the person to whom the message was sent to guess at the meaning, even without a verbal explanation. The following was the method of the Wotjoballuk of the Wimmera river in Victoria.[3] “The messenger carried the message-stick in a net bag, and on arriving at the camp to which he was sent, he handed it to the headman at some place apart from the others, saying to him, ‘So-and-so sent you this,’ and he then gives his message, referring as he does so to the notches on the message-stick; and if his message requires it, also enumerates the days or stages, as the case may be,” by a method of counting on different parts of the body.

For the purposes of communication with absent persons, however, another method commended itself, which in time was Marked pebbles. adopted also for mnemonic purposes. This method was the beginning whence some forms at least of later writing have been derived. From the very earliest times to which the energy of man can be traced, date two kinds of writing: (a) engraving of a visible object on some hard substance, such as the flat surface of a bone; (b) drawing, painting or engraving marks which could again be identified. Of the first kind are the engravings of reindeer, buffaloes and other animals by the cave men of prehistoric times; of the second are a large number of pebbles discovered by M. Ed. Piette at Mas d'Azil, on the left bank of the Arize, an account of which was published by the discoverer in L'Anthropologie (1896), vii. 384 sqq. This layer of coloured pebbles is intercalated between the last layer of the Reindeer Age and the first of the Neolithic period. The layer is over 2 ft. thick, of a reddish-black colour, and along with the pebbles are found cinders, peroxide of iron, teeth of deer perforated, probably in order to be strung like beads, harpoons of various kinds, and the bones of a large number of animals, some wheat, and, in the upper part of the layer, nuts, cherry-stones and plums. The stones were coloured with peroxide of iron. The characters are of two kinds: (a) a series of strokes which possibly indicate numbers, (b) graphic symbols. The stones were scattered about without connexion or relation one with another. Whatever the meaning may be, it is clear that the markings are not accidental. It is noticeable, however, that none of them definitely represent any animal, though some of them bear a certain resemblance to caterpillars or serpents. Others look like rough attempts to represent trees and river plants. A great number closely resemble symbols of the alphabet. Piette himself was inclined to see in the symbols the forerunners of the later syllabaries and alphabets of the East, nine of them agreeing with forms in the Cypriot syllabary (see below) and eleven with those of the Phoenician alphabet. A certain amount of likeness, however, could not well be avoided, for as soon as the artist advances beyond the single perpendicular or horizontal line he must, by crossing two lines, get forms which resemble alphabetical symbols. It might be therefore a safer conclusion to suppose that if they passed beyond magic symbols, to be buried like the Australian churinga, they were conventional marks understood by the members of the clan or tribe which frequented the caves of Mas d'Azil. It has been suggested that, like similar things among the American Indians, they may have been used in playing games or gambling.

A very large number of conventional marks, however, are demonstrably reductions from still older forms, conventional American picture-writing, &c. marks often developing out of pictographs. Pictography has, in fact, left its traces in all parts of the world. It has, however, been most widely developed in the New World as a system lasting down to modern times. The American Indians, besides picture-writing, used also (1) the simple mnemonic of a notched stick to record various incidents, such as the number of days spent on an expedition, the number of enemies slain and the like; (2) wampum belts, consisting of strung beads, which could be utilized as a mnemonic, exactly like a rosary. Wampum belts, however, were employed in more intricate forms; white beads indicated peace, purple or violet meant war. Sometimes a pattern was made in the belt with beads of a different colour, as in the belt presented to William Penn on the making of a treaty with the Leni-Lenape chiefs in 1682. Here, in the centre of the belt, two figures, intended to represent Penn and an Indian, join hands, thus clearly indicating a treaty. Very simple pictures are drawn upon birch bark, indicating by their order the subjects in a series of song-chants with sufficient precision to enable the singer to recall the theme of each in his recitation. An account can be kept of sales or purchases by representing in perpendicular strokes the number of items, and adding at the end of each series a picture of the animal or object to which the particular series refers. Thus three strokes followed by the picture of a deer indicate that the hunter has brought three deer for sale. A conventional symbol (a circle with a line across it) is used to indicate a dollar, a cross represents ten cents, and an upright stroke one cent, so that the price can be quite clearly set forth. This practice is followed in many other parts of the world. In clay tablets discovered by Dr Arthur Evans during his exploration of the great palace at Knossos, in Crete, a somewhat similar method of enumeration is followed; while at Athens conventional symbols were used to distinguish drachmae and obols upon the revenue records, of which considerable fragments are still preserved.

In comparatively recent times, according to Colonel Mallery (10th Annual Report of American Bureau of Ethnology), the Dakota Indians invented a chronological table, or winter count, wherein each year is recorded by a picture of some important event which befell during that year. In these pictures a considerable amount of symbolism was necessary. A black upright stroke indicates that a Dakota Indian was killed, a rough outline of the head and body spotted with blotches indicates that in the year thus indicated the tribe suffered from smallpox. Sometimes, in referring to persons, the symbol is of the nature of a rebus. Thus, Red Coat, an Indian chief, was killed in the winter of 1807-1808; this fact is recorded by a picture of a red coat with two arrows piercing it and blood dripping. There is, however, nothing of the nature of a play upon words intended, and even when General Manyadier is represented as a figure in European dress, with the heads of two deer behind his head and connected with his mouth, no rebus was intended (many a deer), but the Indians supposed that his name really meant this, like their own names Big Crow, Little Beaver, and so forth. Here the Mexicans proceeded a stage further, as in the often quoted case of the name of Itz-coatl, literally knife-snake, which is ordinarily represented by a reptile (coatl) with a number of knives (itz) projecting from its back. It is, however, also found divided into three words, itz-co-atl—knife-pot-water—and represented by a different picture accordingly. The Mexicans, moreover, to indicate that the picture was a proper name, drew the upper part of the human figure below the symbol, and joined them by a line, a practice adopted also amongst their northern neighbours when, as in names like Little-Ring, the representation would hardly be sufficiently definite. Simple abstract notions could also be expressed in this picture-writing. Starvation or famine was graphically represented by a human figure with the ribs showing prominently. A noose amongst the Mexicans was the symbol for robbery, though more logically belonging to its punishment. In a Californian rock-painting reproduced by Mallery (p. 638), sorrow is represented by a figure from whose eyes drop tears. This could be abbreviated to an eye with tears falling from it, a form recorded by Schoolcraft as existing amongst the Ojibwa Indians. The symbol is so obvious that it is found with the same value among Egyptian hieroglyphics.

The civilization of the American Indians was nowhere very high, and for their simple needs this system, without further development, sufficed. It was different in the more elaborate civilizations which prevailed among the ancient peoples of the Old World, to whom with certainty the development of writing from pictography can be ascribed—the Assyrians (see Cuneiform), Egyptians (see Egypt) and Chinese (see China). Here more complex notions had to be expressed. The development of the system can be traced through many centuries, and, as might be expected, this development shows a tendency to conventionalize the pictorial symbols employed. Out of conventionalized forms develop (a) syllabaries, (b) alphabets. As regards the latter the historical evolution is traced in the article Alphabet. The account given under China (language) gives a good idea of the development of a syllabary from pictographic writing.

The Egyptian system of writing is perhaps the oldest of known scripts, and was carried on till the Ptolemaic period, when the more Egyptian. convenient Greek alphabet led to its gradual disuse. But, as in Chinese, the fact that it was so long in use led to the conventionalizing of the pictures, and in many cases to a complete divorcement between the symbol and the sound represented, the original word having often become obsolete. In this case it is no longer possible to trace it. Attempts have been made to connect the three great pictographic systems of the Old World, some authorities holding that the Chinese migrated eastwards from Babylonia, while others contend that the civilization of Egypt sprang originally from the valley of the Euphrates, and that the ancient Egyptians were of the same stock as the Somali and were overlaid and permeated by a Semitic conquest and civilization. But there is no clear evidence that the Egyptian system of writing was not a development in the Nile Valley itself, or that it was either the descendant or the parent of the pictographic system which developed into the cuneiform of Assyria and neighbouring lands.

Egyptian started from the same point as every other pictographic system—the representation of the object or the concrete expression of the idea. But, like the Chinese, it took the further step, short of which the American Indian pictographs stopped; it converted its pictures into a syllabary from which there was an imperfect development towards an alphabet. Egyptian, however, never became alphabetic in the sense in which the western languages of modern Europe are alphabetic. This is attributed to the natural conservatism of the people, and the influence of the artist scribes, who, as Mr F. Ll. Griffith has pointed out, “fully appreciated the effect of decorative writing; to have limited their choice of signs by alphabetic signs would have constituted a serious loss to that highly important body.” The effect of this love for decoration, combined with a desire for precision, is shown by the repetition several times over in the symbols of the sounds contained in a word. The development of Egyptian was exactly parallel to Chinese. A combination of sounds, which was originally the name of an object, was represented by the picture of that object. This picture again, like Chinese, and like the Indian name “Little-Ring,” required at the end a determinative—a picture of the kind of object intended—in order to avoid ambiguity. As the alphabet represented only consonants and semi-consonants, and the Egyptian roots consisted mostly of only three letters, the parallelism with Chinese is remarkably close.

The cuneiform script spread to other people who spoke tongues in no way akin to those of either its inventors, the Sumerians, or their Hittite. conquerors, the Semitic Babylonians. A widespread series of inscriptions, found in many parts of Asia and even in the Aegean, which are generally described as Hittite (q.v.) are written in a script of pictographic origin, though probably independent of Babylonian in its development.

It is noteworthy that at a very early period a colony of Greeks from the Peloponnese, speaking a dialect closely akin to the Arcadian Cyprian. dialect (which is known to us only from a much later period), had settled in the island of Cyprus. Alone among the Greeks this colony did not write in an alphabet, but under some Asiatic influence adopted a syllabary. Even when the island came again closely in touch with their Greek kinsfolk, after the Persian wars, the Greek inhabitants continued to write in their syllabary. In the recent excavations made by the authorities of the British Museum an inscription of the 4th century B.C. was discovered, whereon a dedication to Demeter and Persephone was given first in Greek letters and repeated below in the syllabary, the Greek (as universally at so late a period) reading from left to right, the syllabary from right to left. This syllabary has five vowel symbols, but it could not distinguish between long and short vowels. In its consonant system it is unable to distinguish between breathed, voiced and aspirated stops, thus having but one symbol to represent τε, δε and θε. It is, of course, unable to represent a final consonant, but this is achieved by using the symbol for a syllable ending in e conventionally for the final consonant. Thus ka-se stands for κάς, the Cyprian equivalent of καί, “and.” There are symbols for ta, for te, for ti, for to, for tu, though none for t, and similarly for most of the other consonants. There is, however, no symbol for wu (Fv); ya, ye, yi occur, but no yo or yu. Δημήτρι is expressed by ta-ma-ti-ri, where ti stands for t alone; sa-ta-sa-to-ro stands for Στασάνδρω (genitive). Here it is to be observed (1) that ν preceding another consonant is omitted altogether, the vowel being probably nasalized as in French; (2) that, as in the previous word, there is a sort of vowel euphony whereby the unnecessary vowel accompanying t takes the colour of the succeeding vowel. In other cases, however, it follows the preceding vowel, as in a-ri-si-to-pa-to-o-a-ri-si-ta-go-ra-u = Ἀριστόφαντο(ς) ὁ Ἀρισταγόρου.

For literature on the history of writing, see the bibliographies to the articles Alphabet, &c., and under the headings of the various languages and peoples.


  1. Quoted by Middendorf, Das Runa Simi oder die Keshua Sprache (Leipzig, 1890), p. 8.
  2. Cf. Andree, Ethnologische Parallelen and Vergleiche, i. p. 184 sqq.
  3. A. W. Howitt in Journal of the Anthropological Institute, xviii. (1889), p. 318 sqq.