Page images

may have been produced. I cannot see it myself. The differences between these three branches, and between them and other languages, have continued so long that we can hardly suppose them to have had a common origin. Let us take, for example, two of the most ancient languages, the Chinese and the Sanskrit. These two languages, or as the latter is not now spoken, the languages derived therefrom, have been proved to have existed at least 4000 years ; the one monosyllabic and atactic, the other polysyllabic and syntactic.* It does not appear that during this period they have at all approached nearer to each other, and, in their general structure and character they remain the same; and if such be the case, are we to assume that prior to the time to which we are enabled to trace them, language could have undergone so many changes in its organic structure as to have produced the different systems of Asia and Europe. But if we are to take the vulgar theory of the age of the world—say 6000 years—to be the correct one, the thing is still more improbable ; for if there has been no material change in these two languages during the last 4000 years, on what ground are we to presume that some violent change should have taken place within the previous 2000 years. The same remarks might be made on other languages, as the Hebrew and kindred languages; the Basque, the Greek, and the Slavonic languages. The Tatar conquest made no alteration in the structure of the Chinese idiom; nor has the Basque lost its grammatical forms, notwithstanding that the people of this part of Spain have been for ages surrounded by nations speaking languages whose idioms are entirely opposite.t

Max Müller makes no mention of the African and American languages, or the Polynesian and Australian dialects, which, we shall hereafter see, amount in number to nearly all the languages of the globe put together. All these languages and dialects are more or less ancient, and some of them may be traced, it is said, as far back

• It is a remarkable fact that in Japan two languages exist at the same time; the one monosyllabic and atactic, like the Chinese; the other polysyllabic, with numerous inflexions and grammatical forms. The former is called the Koye; the latter the Yomi. Both are in use at the same time, and occasionally intermix with each other, still preserving their general character and peculiar structure. See Elémens de la Grammaire Japonaise, par le P. Rodriguez, traduits du Portugais, par M. C. Landusse, Paris, 1825; and Encyc. Amer. (Lieber); Phil., 1832, in voc. “ Philology".

+ Cf. Partington, Brit. Cyc. Between the Chinese and the Cherokee it would be difficult to find the least etymological affinity; and if the distance of places is assigned as the cause, we will instance the Bengáli, a language poken in a country not far from China, and which differs from the Chinese full as much as Mohawk from Potawotamee; ib.

as the Chinese and Sanskrit; and yet their organic differences have remained the same for ages. We find in them idioms of different structure, having characters of their own, of which it would be in vain to seek for traces in a primitive tongue.* Independently, however, of grammar, a comparison of words should be made. That the European and Asiatic languages have many words in common there can be no doubt. On the other hand, some of the most ordinary words are totally dissimilar in very many languages; and I am disposed to think, that taking into account phonetic decay, wear and tear of words, and other causes, there remains scarcely any possibility of their having had a common origin. I will merely give a few of such words, reserving for a future paper to extend the list. The word gold, which in the Gotho-Teutonic languages is found written gull, guld, and goud, and in the Tatar goltz, can have no etymological connection with Latin aurum (whence Spanish and Italian oro, French or, Gaelic dir); nor with Polish zloto, Greek Xpvoeos, Sanskit kănă , Arabic zahab, Persian zur and tilá, Hindústání and Bengali sond, Turkish altún, Malay amas or mas, Quichua cúri, Bugis ulawong, Egyptian vouß, Anamitic vàng, Chinese kin, Malagasy volamena. Compare English moon, Latin luna, Greek oeAnun, Persian parú, Turkish âï, Sanskrit chandra; English water (Greek vòwp), Latin aqua, Malagasy mandena, Birmese re, Mandchu mouke, Bugis uwae (Oceanic vai), Tonquinese nou-di, Chinese shůhy; English sea, Arabic bahr, Turkish dengiz, Sandwichian kai; English house, Greek oukos, Mandchu po, Arabic bait (Hebrew beth), Marquisian hae, fae, Chinese ůh, fang üh, kea, chou kea, Turkish av, Bugis bolah, Hin-" dústání ghar; English mountain (Latin mons), German berg, Greek opos, Turkish tágh, Pushto ghar, Quichua urcu; English bread, Latin panis, Finnish leipä, Malay róti Chinese mëen tow, mëen paou, Javanese redjekki, Japanese moci, Armenian zhats, Georgian puri, Coptic oik, Ethiopic sifai, Chilian cobque, Mexican remiou.

I will conclude with a few remarks upon the statistics of language. Francesco Lopez, a native of South America, who had extensive knowledge of both continents, thought it no rash statement to make, that the idioms, notabilmente diversi, of both Americas amounted to at least 1,500.† The Abbé Rozo says that the inhabitants of the two Americas spoke not less than 2,000 languages. Father Kircher informs us that American missionaries make the South American languages amount to 500; while the Abbé Clavigero had cognizance

[ocr errors]

• Partington, Cyc.

+ See Hervas, Cat. Ling., p. 11.

of 35 idioms spoken in Mexico. The Imperial Dictionary,* compiled at the instance of the Empress Catherine, which was published in the year 1787, contains a list of 285 words, translated into 51 European, and 149 Asiatic languages. A second edition of this work, in which the words are arranged alphabetically, appeared in 1790-91, in 4 vols., edited by Jankiewitsch de Miriemo. This edition contains, according to some, 279 languages; 171 for Asia, 55 for Europe, 30 for Africa, and 23 for America. The authors of the Mithridates increased the number of known languages and dialects to 2,000; which Friedrich Adelungt augmented to 3,066, geographically distributed as follows:Asiatic

987 European

587 African

276 American


3,0649 The numbers of those who speak the different languages made use of in America are thus distributed. English, 11,647,000; Spanish, 10,174,000; Portuguese, 3,740,000; Indian, 7,593,000; French, 1,242,000; Dutch, Danish, Swedish, and Russian, 216,000.||

With respect to the number of words contained in some of the principal languages, in the following, I can only give the approximate number. The Arabic contains about 150,000; the Hindústání, 18,000; the Sanskrit, 27,000; the Malay, 13,000;1 the Puk'hto or Pus'hto, 22,000;** the Egyptian, 4,000; the Armenian, 30,000; the Turkish, 50,000; the Mandchu, 16,000; the Latin,

* Glossarium Comparativum Linguarum totius Orbis. Petersb., 1787.

+ According to Pott (Ungleichheit, p. 230), it contains 277 languages ; 185 for Asia, 22 for Europe, 28 for Africa, 15 for America. This would make 280. Max Müller. [It would rather add up 250. R. S. C.]

| Ubersicht aller bekannten Sprachen und ihrer Dialekte, von Friedrich Adel. ung, 8vo, pp. xiv 186. St. Petersb., 1820.

$ About twenty of the Italian dialects have been reduced to writing, and made known to the press. The dialects of France are almost as numerous as her pro. vinces. Languedoc alone has seven or eight distinct dialects.-Champollion-Figeac reckons the most distinguishable dialects of France at fourteen. The number of inodern Greek dialects is carried by some as high as seventy. (Cf. Marsh, p. 678; Sir John Stoddart's Glossology, s. 31, and p. 29 and 33.) The principal British dialects are those of Norfolk and Suffolk, Kent, Durbam, Gloucester. shire, Essex, Yorkshire, Worcestershire, Herefordshire, Dorset, Sussex, Devonshire, Warwickshire, Cornwall, Derbyshire, Westmoreland, Cumberland, Sheffield, Shropshire, Lancashire, Somersetshire, Chesbire, Northumberland, Craven, and the Scottish dialect. || Encyc. Amer. (Lieber); Phil., 1829.

Marsden's Dictionary gives about 6,000; Crawfurd's about 13,000; which includes many proper names.

** Larramendi, in his day, gave the number at 13,365 only.


40,000; the mediæval Latin, 100,000; the Greek, 89,000; the Spanish, 25,000; the Italian, 50,000; the French, 42,000; the Gaelic, 23,000; the Irish, 50,000; the Welsh, 40,000; the Russian, 40,000; the Polish, 24,000; the Anglo-Saxon, 25,000. Flügel, in 1843, estimated the number of German words in his own dictionary at 94,464, of which 65,085 were simple, 29,379 compound. Thommerel gave the number of words in the English dictionaries of Robertson and Webster as 43,566, 29,853 of which he derives from Classical, 13,230 from Teutonic (Anglo-Saxon), and the rest from miscellaneous

Todd's edition of Johnson, however, is said to contain 58,000 words, and the later editions of Webster, which contain the particles of the present and perfect, have reached 70,000 ;* but, if every word were included, the number would probably now exceed 80,000. The Hebrew words in the Old Testament amount to 5,643. The Hindí has exactly 6,000 words ;f the languages of the Marquesas and the Sandwich Islands, 6,123; the Provençal, 107,201. The cuneiform inscriptions of Persia contain no more than 379 words, of which 131 are proper names. The vocabulary of the ancient sages of Egypt, as far as it is known to us from the hieroglyphic inscriptions (as given by Bunsen), amounts to about 658 words. The number of hieroglyphic groups in Sharpe's Egyptian Hieroglyphics (1861), amounted to 2,030. There are about 450 radicals or sounds in the Chinese language, which by various accents and intonations are raised to 1,263. I Mr. Crawfurd says an examination of 4,074 radical words of the dictionary shows that the Malay language is composed of the following lingual elements :-Native Malay words, 2,003; common to the Malay and Javanese, 1,040; Sanskrit, 199; Telinga, 23; Arabic, 160; Persian, 30; Portuguese, 19. He says, further, an examination of the Malay, including its foreign elements, shows that, out of 1,000 words, 285 are common to it and the Javanese; and a similar

• Cf. Marsh's Lectures, p. 182; and Max Müller, p. 27). + Cf. Dr. Hunter's Hindústání Dictionary.

The exact number of words contained in the Imperial Dictionary of Khanghi amounts to 42,718. About one-fourth part has become obsolete, and one-half of the rest may be considered of rare occurrence, thus leaving only about 15,000 words in actual use. The number of the classical characters is 42,718, but many of them are no longer in use in the modern language. (Stanislas Julien.) CF. Müller.

$ Grammar and Dictionary of the Malay Language, by John Crawfurd, F.R.S. 8vo, London, 1852. At vol. i, p. 73, speaking of the whole body of the language, he says it contains 516

words from the Sanskrit ; 750 from the Arabic; 95 from the Persian ; 40 from Tâlugu or Telinga ; and 37 from Portuguese. He says, further, that the earliest example of Malay is the Vocabulary of Pigafetta, in 1521, which contains 344 words only, 270 of which can be readily ascertained to be the same language as that spoken at the present day.



one of the Javanese, that 240 out of 1,000 are common to it and the Malay; and that of the Malay 715 parts, and of the Javanese 760, appear to be native.

Professor Müller tells us that Sanskrit grammarians have reduced the whole growth of their language to 1,706 roots ;* but he is of opinion that the primitive sounds expressive of different meanings requisite for the etymological analysis of the whole Sanskrit dictionary would not amount to one-third of that number, and he doubts whether they may not be reduced even to 500 words; that Renant has reduced the Hebrew to about the same number; and that Benloew! estimates the necessary radicals of Gothic at 600, and of modern German at 250. The Latin primitives contained in the Index Etymologicus of Gesner's Thesaurus, amount to 2,400;$ but, as I have before said, they may be reduced to some 900 or 1,000 words. The Greek primitives given by MM. Port Royal amount to 2,200, might, perhaps, be reduced to about 1,200.

The following tabular form gives the proportions of vowels and consonants in some of the principal languages. [ Sandwich Islands



1 1.333 .

1 1.006 Portuguese

1.02 1 Common Arabic


1* Italian

1.1 1 Seneca Indians

1:18 1 Chahta Indians

1.2 1 Sanskrit


1* Latin


1 Hebrew


1* Spanish

1.24 1 Persian

1.33 1 Malay

1.33 1 French, phonic prop.

1:34 1 orthographic 1.27: 1 Dutch

1.5 1 English, phonic prop.


orthographic 1.52 : 1 Swedish

1 German, phonic prop.


1.7 1 orthographic 1.64 : 1 Benfey, Grammatik, s 147. + Histoire des Langues sémitiques, p. 138. P. 22. § See Vans Kennedy.

|| Ibid. See Encyc. Amer. (Lieber), Phil. ; 1830, in voc. “ Consonants". Those marked with * are counted phonically.

Greek { Attic dial.

[ocr errors]

« PreviousContinue »