Arrangement.— The first condition of use fulness being that the items shall be easily found, no arrangement is tolerable except an alphabetical one, in all languages where the in flections are mainly terminal or after a conso nant initial. This arrangement, however, may be used only in a key, making feasible other classifications for the main body; as in Roget's well-known 'Thesaurus' of synonyms, where the grouping is by ideas, in parallel column with the reverse ideas. This would be a very useful addition to the meagre lists of synonyms given in the large dictionaries; but it would need great labor and almost impossible accuracy of cross-referencing, as the group heads must be referred to from each word in its alphabetical place. A very useful variant is in the lists of technical terms of arts and sciences given under each main head in the 'Standard Dictionary.) Rhyming dictionaries are arranged alphabeti cally under the terminations. In etymological dictionaries, attempts have been made to alpha betize by roots; but as no two scholars ever agree on the roots, the finding is largely guess work, unless with a full alphabetical key, and scholars condemn the plan. In Semitic lan guages the arrangement is usually by roots. It was this necessity of arranging Volapiik dic tionaries, owing to the prefix-inflection involv ing a search in three or four places for one word, which as much as anything else prevented that (language' gaining favor. The arrange ment of Chinese characters is almost incapable of any readily intelligible system, and is not set tled by sinologists.
History.— The first dictionaries were used by the Assyrians and Babylonians, to explain not words, but signs. The ancient ideographs, when largely displaced by syllabified or alphabetic signs with phonetic values, rapidly became un intelligible; and syllabaries were therefore com piled, to define them in terms of the latter. The old Sumerian language, however, seems not then to have become adead,D being preserved by the colleges of priests for religious use; and foreign languages were learned from foreigners. These were inscribed in vertical columns, on clay tab lets; they have been found in the great library of Asurbanipal (Sardanapalus —668-626 B.c.), at Nineveh„the source of our chief knowledge of Mesopotamian culture. Much the same sort of syllabaries seem to have been used by the Chinese and Japanese.
The ancestry of our own dictionaries, how ever, is Greek. Here we observe the usual pro gression from the special to the general. The typical modern dictionary is a complete or rep resentative vocabulary of a language, present and historical, with definitions in its own or some foreign tongue; its object being to inter pret either the meaning of the foreign tongue or the bygone part of its own, or the usages and history of the latter. But neither of these was thought of by the ancients. The small upper and priestly class settled their own usages of language, there was no half-educated reading populace relying on books for correct form, there was no foreign literature they cared to read, and foreign conversation they learned from con versation. Their first dictionaries were the
same as our special vocabularies. The oldest surviving one, by Apollonius of Alexandria in Augustus' time, was a glossary of Homer's words—probably the oldest of all kinds if we had them, Homer being the common textbook of Greece; others were of ambiguous, corrupt, barbarous, foreign or dialectic phrases in the tragic and comic poets—phrases Attic, Lace demonian, Cretan, Rhodian, Italian, Phrygian, etc.; others of technical terms in cookery, a fas cinating subject to the Greeks; others of driW ing vessels, of the cries of animals, of syno nyms, etc. Besides glossaries of Homer, there were others of Plato, the Ten Attic Orators, Hippocrates, etc. All of these, so far as pre served — unhappily, very few of them arc pre served, and Atheneus in the 2d century gives the names of 35 lost to us—are of priceless value both for words and antiquities. On a larger scale were the Onomasticon of Julius Pollux, arranged topically in 10 books, which has survived; and a great lexi con, now lost, written by Helladius of Alexan dria about 400 u.c. Of others may be noted an etymological dictionary by Orion of Thebes about 450, the first we have— for Varro's es says in Roman etymology can hardly be called a dictionary; a dialectic and local one by Hesychius of Alexandria, in the 4th century; and one of words similar in sound, but unlike in meaning, by Ammonius of Alexandria. Com ing down to the Middle Ages, the most famous is that of Suidas, of unknown date or place, but probably about the 10th century; it is a slovenly compilation of words, names and places, but 'extremely valuable from its matter, not to be had elsewhere. Overleaping some centuries and their works, we come in 1572 to the mightiest of all, the tremendous Graeae Linguae) of the famous French scholar and printer, Henri Estienne or Stephanus (q.v.), Geneva, 4 vols., folio, a work gigantic not only in size but in scholarship • it was last re printed by Didot at Paris, 1831-65, ed. Din dorf, 9 vols., folio, 9,902 pages. All the medie val and modern lexicons till the 19th century were Greek into Latin; the first Greek-English one was not till 1814, by John Pickering, and that was not finished; the first complete one was by John Jones, 1823, London. For many years the only general one on the market has been the Liddell & Scott, revised by Drisler, in its various editions. This, however, is confined to classic and early-empire forms, to common nouns and to book words; and leaves need for several ad ditional words. Later and medieval Greek has been admirably covered by the lexicon of E. A. Sophocles, revised 1::7 by J. H. Thayer; and a three-volume lexicon entirely devoted to the vast wealth of forms furnished by the inscrip tions in progress. There is one of modern Greek by Contopoulos, Athens 1889; and there are special glossaries of the New Testament and of various great authors, English-Greek, Yonge, revised by Drisler, latest edition 1893.