182 DICTIONARY and complete explanations of all the meanings, and again, as is necessary in a complete dictionary, may include usage. The explanations of the meanings should be precise and not vague, real definitions and not a mere reference of one word to another of the same meaning, as when the French Academy explains fier by hautain, altier, and hautain by jier, orgueilleux. But when one language is explained by another, nothing conveys the meaning so well as a perfectly equivalent word. The interpretation of a language by itself is, as Dr Johnson says, very difficult, for there is no other word to express the idea, and simple ideas cannot be described. Therefore, in Grimm s dictionary Latin and other languages are used when necessary. Synonyms and homonyms should be given, as well as words of opposite meaning, and their similarities and differences explained. Remarks should be made on difficulties, faults to be avoided, peculiar constructions, figurative, idiomatic, and proverbial expressions, and the origin of these given when possible. All this should be done in the fewest and plainest words. Eloquence is out of place in a dictionary; bat the author must not fear fulness when it is necessary, and must not allow brevity to make him obscure. A com plete dictionary of a copious language must necessarily be a very large book, but much space may be saved by the use of well-selected terms and abbreviations, and by typo graphical arrangements. Examples form a very important part of a dictionary, but one which is generally omitted, often neglected, and seldom so carefully attended to as it deserves. When no quotations are given, the whole lauguage depends on the authority of the author of the dictionary. The French Academy have always claimed the right of making their own examples. Voltaire says they seem to have made a law not to quote, but, he acids, a dictionary without quotations is a skeleton. Examples may be arranged either under the meanings they illustrate, which is the usual and most useful plan, or, in languages possessing an extensive literature of long duration, chrono logically in one series, as the Philological Society formerly proposed. Littre" has adopted a medium, and gives examples from authors of the 17th, 18th, and 19th centuries under the meanings to which they belong, and those from previous authors in a chronological series. Each quotation should give a complete sense, and not be a mere fragment of a sentence. It should, if possible, be instructive and interesting in itself, but should not on this account be made too long. Those containing etymologies, definitions, or explanations of a word, as well as those in which it is joined to words of the same or opposite meaning, and those which mark its intro duction or disuse, and those in which it is used as a foreign word not yet naturalized, should be especially sought for. Each should have as exact a reference as pos sible. The common practice of giving only the author s name makes it sometimes impossible to verify a quotation without searching through his entire works, which may fill many volumes. In the case of some rare words, when the quotation would add nothing to the information other wise given, the mere reference may suffice. The value of a dictionary and the richness of its vocabulary depend very much on the carefulness and extent of the search for examples, which can only be complete when it has ex tended to the whole literature of the language. If con cordances and full indexes were more universal, the search for examples would be much facilitated. The foundation of the Philological Society s intended dictionary was to have been the reading of all English books not purely scientific for examples by volunteers. In October 18G4, 1149 had been read, and 360 were in hand. Though complete dictionaries of a language are very few, and none as yet exists in English, large dictionaries are many. The tendency of great dictionaries is to unite in themselves all the peculiar features of special dictionaries. A large dictionary is most useful when a word is to be thoroughly studied, or when there is difficulty in making out the meaning of a word or phrase. Special diction aries are more useful for special purposes ; for instance, synonyms are best studied in a dictionary of synonyms. And small dictionaries are more convenient for frequent use as in translating from an unfamiliar language, for words may be found more quickly, and they present the words and their meanings in a concentrated and compact form, instead of being scattered over a large space, and separated by other matter. Dictionaries of several languages, called polyglots, are of different kinds. Some are polyglot in the vocabulary, but not in the explanation, like Johnson s dictionary of Persian and Arabic explained in English ; some in the interpretation, but not in the vocabulary or explanation, like Calepini Octoglotton, a Latin dictionary of Latin, with the meanings in seven languages. Many great dictionaries are now polyglot in this sense. Some are polyglot in the vocabulary and interpretation, but are explained in one language, like JaPs Glossaire Nautique, a glossary of sea terms in many languages, giving the equi valents of each word in the other languages, but the ex planation in French. Pauthier s Annamese Dictionary is polyglot in a peculiar way. It gives the Chinese charac ters with their pronunciation in Chinese and Anuamese. Special dictionaries are various, and many kinds will be found in the following list. There are dictionaries of etymology, foreign words, dialects, secret languages, slang, neology, barbarous words, faults of expression, choice words, prosody, pronunciation, spelling, orators, poets, law, music, proper names, particular authors, nouns, verbs, participles, particles, double forms, difficulties, and many others. Fick s dictionary (Gottingen, 1SGS. 8vo ; 1 874-7 G, 8vo, 4 vols.) is a remarkable attempt to ascertain the common language of the Indo-European nations before each of their great separations. In the second edition of his Etymologische Forschunyen (Lemgo and Detmoldt, 1859-73, 8vo, 7217 pages) Pott gives a comparative lexicon of Indo-European roots, 2226 in number, occupy ing 5140 pages. Comparatively few languages possess dictionaries, and they are few in number compared to other books, probably much under 2 per cent.; and 5000, not counting different editions, might be considered a very large collection. More than half belong to European languages, of which five surpass the rest in the number and variety of their dictionaries, namely, Greek, Latin, French, English, and German. In Asia, those excelling in this respect are Hebrew, Arabic, Persian, Sanskrit, Hindustani, Malay, Chinese, and Japanese ; in Africa, Egyptian, Ethiopic, and Kafi re ; in America, Otoini, Aztec, Guarani, Tupi, and Quichua. The following list of dictionaries is arranged geographi cally by families of languages, or by regions. In each group the order, when not alphabetical, is usually from north to south, extinct languages generally coming first, and dialects being placed under their language. Diction aries forming parts of other works, such as travels, histories, transactions, periodicals, reading-books, &c., are generally excluded. When a selection has to be made, the earliest, largest, latest, and best dictionaries are preferred. This system, seemed on the whole best calculated to keep together dictionaries naturally associated. The languages to be considered are too many for an alphabetical arrange ment, which ignores all relations both natural and geo graphical, and too few to require a strict classification by affinities, by which the European languages, which for
many reasons should be kept together, would be dispersed.