LANGUAGE. 



groups. In these, the relational part of thought 

 obtains prominent vocal expression by separate 

 roots, joined or glued on to the significant roots as 

 terminations. These terminations were originally 

 themselves significant roots, and many of them 

 are still used as separate significant words, al- 

 though the greater part have sunk down to mere 

 signs of cases and other relations. The compound 

 expression thus formed never, however, attains 

 perfect unity ; the significant root always remains 

 rigid, unobscured in its sense, and unchanged in 

 form, and the termination is felt as something 

 distinct from the body of the word. 



Thus, the Finnish declension exhibits a struc- 

 ture of the most mechanical and transparent kind 

 e.g. karhu, bear ; karhu-n, of the bear ; karhut- 

 ta, without bear ; karhu-sta, out of the bear ; and 

 so on through fifteen cases. The insertion of the 

 plural suffix, i, gives karhu-i-n, of the bears ; 

 karhu-i-ta, without bears ; karhu-i-sta, out of the 

 bears ; &c. But this composite mechanical struc- 

 ture reaches its climax remaining all the while 

 perfectly transparent in the Turkish verb. Thus, 

 the root sev has the indefinite meaning of loving, 

 and the inf. is sev-mek, to love ; which then, by 

 the insertion of certain suffixes, can take on as 

 many as forty forms or voices e. g. sev-me-mek, 

 not to love ; sev-e-me-mek, not to be able to love ; 

 sev-dir-mek, to cause to love ; sev-dtr-ish-mek, to 

 cause one another to love ; sev-il-mek, to be loved ; 

 sev-il-e-me-mek, not to be able to be loved ; &c. 

 Each of these forms, then, runs through a large 

 round of tenses and moods, with their persons and 

 numbers. 



The languages of the American Indians are all 

 of this agglutinating type, although they have 

 also got the name Incorporative, or Intercalative, 

 because they run a whole phrase or sentence into 

 one word e. g. hoponi, to wash ; hopocuni, to 

 wash hands ; hopoaduni, to wash feet ; ninacaqua, 

 I (nt) eat (qua] flesh (naca). The Basque lan- 

 guage partakes of this character. In building 

 these compounds, it mostly happens that only 

 fragments of the single words are used. Thus, in 

 Mexican, alt, 'water;' chichiltic, 'red;' tlacatl, 

 'man;' and chorea, 'weep,' are clipped and fused 

 into achichillachocan, which means, 'the place 

 where people weep because the water is red.' 

 Similarly, the Basque language makes ilhun, f twi- 

 light,' out of hill, 'dead/ and egun, 'day;' and 

 belhaun, ' the knee,' from belhar, ' front,' and oin, 

 'leg.' 



It is only in the third or Inflectional stage that 

 perfect unity of the two elements is attained. In 

 the Aryan and Semitic tongues, which alone have 

 reached this highest state of development, the 

 significant root and the termination have become 

 blended into one both in effect and form, and 

 phonetic changes have for the most part obliter- 

 ated the traces of composition. Yet no doubt is 

 felt by philologists that the most highly organised 

 of the inflecting or amalgamating languages began 

 with the radical stage, and passed through the 

 agglutinate. The analytic powers of comparative 

 grammar have succeeded in tracing back the 

 formal elements of the Aryan tongues to original 

 independent words, agglutinated to other words 

 to modify them. Of this we have given 

 numerous examples when speaking of word- 

 building. Against this theory it has been 

 urged, that there is no historical instance of a 



language so changing its type, and passing from 

 one stage to another. But a sufficient account of 

 this phenomenon may be found in the different 

 mental habits and political positions of the peoples 

 (see Max Miiller, Lectures on the Science of Lan- 

 guage, First Series, page 316). Besides, the lan- 

 guages of the lower types do shew a tendency, 

 under favourable circumstances, to produce gram- 

 matical forms of the higher kind. Even in 

 Chinese, in some of its modern dialects, something 

 like cases is to be seen ; and Finnish and Turkish, 

 in contact with the inflected languages of Europe, 

 are making approaches to the inflectional type. 



On the other hand, the inflectional languages 

 had, before the earliest times of which we have any 

 written monuments, entered on the reverse phase 

 the analytic. By the process of phonetic change 

 and decay, the grammatical forms have been 

 gradually becoming obliterated and losing their 

 power, and their place has been supplied by 

 separate words, in the shape of prepositions and 

 auxiliary verbs. 



No satisfactory classification of the monosyl- 

 labic and agglutinate languages has yet been 

 made. This is owing partly to the nature of their 

 structure, and partly to the circumstance that they 

 have been less thoroughly studied. They may be 

 arranged in groups having more or fewer points 

 of similarity; but it has not been conclusively 

 shewn with regard to any group that the members 

 of it are genealogically related that the features 

 of resemblance are owing to a common parentage. 

 The more cautious school of philologists object to 

 the term 'Turanian Family,' which has been 

 much used to comprise ' all languages spoken in 

 Asia and Europe (including Oceania), and not 

 included under the Aryan and Semitic families, 

 with the exception of Chinese and its cognate 

 dialects.' They would restrict the term Family to 

 the two genealogical groups, the Aryan and the 

 Semitic. Only in one other case, that of five of 

 the so-called 'Turanian Family,' the Finnic, 

 Samoiedic, Turkic, Mongolic, and Tungusic, 

 would the evidence of family affinity seem to be 

 strong; and even that is spoken of with hesita- 

 tion. This group has been called the Alatyan 

 group or family, from the name applied by the 

 Tatars of Siberia to themselves. As a convenient 

 way of designating the languages which are not 

 Aryan and not Semitic, it has been proposed to 

 call them Sporadic, t.e. scattered, or Allophylian, 

 i.e. spoken by other different tribes (see Farrar's 

 Families of Speech}. 



On the principles of classification above sketched, 

 the chief languages of the earth may be thus 

 arranged : 



I. Monosyllabic or Isolating. i. Chinese, the 

 typical language of this order. 2. Tibetan, which 

 shews some beginnings of grammatical forms. 3. 

 The languages of the Eastern Peninsula Siamese, 

 Anamese, Burman. Japanese and the language of 

 Corea are doubtful 



II. Agglutinate. I. The Alatyan group above 

 described. 2. The Dravidian languages spoken in 

 the south of Hindustan (Canarese, Tamul, Telugu, 

 &c.). 3. North Asiatic (Kamtchatdales, Kuriles, 

 &c.). 4. Malayo- Polynesian group. 5. African 

 Languages. Some of the languages of Africa are 

 allied to the Semitic family, and were introduced 

 by immigration, such as the dialect of Tigrd in 

 Abyssinia and the Arabic dialects spoken by the 



