Page:The American Cyclopædia (1879) Volume X.djvu/155

 LANGUAGE 149 within this group, have been given in the ar- ticle GERMANIC KACES AND LANGUAGES. The roots of the Semitic languages (Hebrew, Chal- dee, Aramaic, Arabic, and others) generally consist of three or more consonants. Chinese roots have generally but a single consonant followed by one or two vowels. Outside of the Indo-European languages little can be defi- nitely established, as it is requisite, in order to attain positive etymological results, that not a single link in the historical connection between a language discussed and the ancient mother language should be wanting. Rei- nisch, in his work, Der einlieitliche Ursprung der Sprachen der alien Welt (Vienna, 1873), has attempted to establish the intrinsic con- catenation of the languages of central Afri- ca, Ethiopia, Egypt, Arabia, Syria, and the Aryan family of speech; but there are still many gaps to be filled up to render the subject entirely clear. F. Lenormant's endeavors, in jfitudes accadiennes (Paris, 1873 et seq.), and in La magie chez les Chaldeens et les origines accadiennes (1874), to develop Jules Oppert's opinion that at the basis of some of the oldest Semitic languages, as Elamitic and Assyrian, lie Finnic and Ugrian (or Turanian) strata of languages, are also far from conclusive. Hew- itt Key, in his recent work mentioned above, has expressed the opinion that the Indo-Eu- ropean languages are closely connected with the speech of the Finns and Lapps; but this opinion also can hardly be considered substan- tiated. Nevertheless these works and sim- ilar ones, as Delitzsch's Studien uber die in- dogermanisch-semitische Wurzelverwandtschaft (Leipsic, 1873), and Conner's VergleicJiendes Worterbuch der Finnisch- Ugrischen Sprachen (Helsingfors, 1874), indicate that many schol- ars perceive that the Indo-European, Semitic, and Turanian groups of languages may possibly have been derived from some one primitive form of speech. The opinions of Lazarus Gei- ger and other theorists on the origin of lan- guage may therefore be finally established by genuine etymological researches. But even if there is hope of an ultimate demonstration of the intrinsic oneness of all human speech, the difficulties still to be overcome are enormous. The phonetic changes which transform words of one language into almost unrecognizable sounds in another, somewhat distantly related, call for most searching examination of the conditions of speech in the various races. The laws of sound are circumscribed by physical conditions. Many languages are entirely de- void of certain sounds ; thus the Chinese can- not produce many European utterances, saying Yamelika for America ; and the aborigines of the Society islands say Tut instead of Cook. Friedrich von Schlegel asserts that the Aztec language has not the sounds of 5, d, /, g, r, , j, v; the Otomi lacks /, i, &, I, r, s; the To- tonaka lacks 5, d, /, r; the negroes have no r, the Australians no s ; most Polynesian lan- guages have no sibilants whatever, and others have only seven consonants, which is the lowest number known. These imperfections and dif- ferentiations of the organs of speech render etymological researches exceedingly difficult. The usual alphabets of from 20 to 26 letters admit of the construction of many billions of words, and these letters are far from sufficient to represent the sounds of all languages. The hopelessness of ever building up the complete laws of the phonetic changes occurring in the almost 6,000 languages and dialects known, is further increased by the fact that our knowl- edge of human speech is confined to historic periods, and that the beginnings of language in prehistoric times, of which no monuments have come down to us, are highly essential to the construction of a satisfactory etymological system. On examining the savage languages now spoken, which many regard as counter- parts of the forms of speech used in the child- hood of mankind, it is found that the simplest sounds often signify the very opposite in other languages of the same degree of development. The sounds most easily produced, 5#, pa, ma, and da, are generally expressions for father and mother ; but what signifies father in one language, signifies mother in another ; thus in ^Georgian mama is father, and dada mother; and in Tuluva, amme father, and appe mother. It has long been evident that the mere compar- ison of words would not be productive of sat- isfactory results. In fact, the day that Bopp first conceived the idea of bringing the test of the method of inflection to bear upon the ques- tion of the affinity and development of tongues, was the real birthday of the science of lan- guage. Grammar is the scientific understand- ing and explanation of the sounds, forms, and functions of words and their parts, and of the construction of sentences. Comparative gram- mar seeks, by comparing the grammars of sev- eral languages, to reach the laws of inflection and construction common to them, and possi- bly to all languages. General or historic gram- mars attempt to explain the growth of language within a specified group of languages. "When languages are analyzed in any state already reached, and not in a state of transition, they become the subject of special grammars, be- longing to the province of linguistics. Com- parative and historical grammars have almost exclusively been written on the Aryan or Indo- European family of speech, enumerated below. It is generally held that the genealogical rela- tion and order of these languages has been demonstrated ; and, though conjecturally only, yet with a tolerable degree of certainty, the extinct and primitive languages spoken by the races before separating into new branches have been reconstructed. Johannes Schmidt, in Die Verwandtschaftsverhaltnisse der indogerma- nischen Sprachen (Weimar, 1872), objects to the idea of a genealogical tree of the Aryan or Indo-European languages, and proposes in its stead a kind of geographical basis of classifica- tion; saying: "You no sooner consign to the