2. From the consideration of the crude materials of lan- guage, we pass on to the varieties exhibited in its struct- ure, with a view to ascertain whether in these there exists any bar to the idea of an original unity. (1.) Reverting to the classification already noticed, we have to observe, in the first place, tliat the principle on whicti it is based is the nature of the connection existing be- tween the predicable and the relational or inflectional elements of a word. ' In the isolating class these two are kept wholly distinct; relational ideas are expressed b}' juxtaposition or by syntactical arrangement, and not by any combination of the roots. In the aggluti- native class the relational elements are attached to the principal or predicable theme by a mechanical kind of junction, the individuality of each being preserved even in the combined state. In the inflecting class the junc-
TONGUES, CONFUSIOX OF 468 TONGUES, CONFUSION OF
tiou is of a more perfect character, and may be com- pared to a chemical combination, the predicable and relational elements being so fused together as to pre- sent tlie appearance of a single and indivisible word. It is clear tliat there exists no insuperable barrier to ori"inal unity in tliese differences, from the simple fact that every iutiecting language must once have been agglutinative, and every agglutinative language once isolating. If tiie predicable and relational elements of an isolating language be linked together, either to the eye or the ear, it is rendered agglutinative ; if the ma- terial and formal parts are pronounced as one word, eliminating, if necessary, the sounds that resist incor- poration, the language becomes iutiecting. (2.) In the .second place, it sliould be noted that these three classes are not separated from each other by any sharp line of demarcation. Not only does each possess, in a measure, the quality predominant in each other, but, moreover, each graduates into its neighbor through its bordering members. The isolating languages are not wholly iso- lating: they avail themselves of certain words as re- lational particles, though these still retain elsewhere their independent character; they also use composite, though not strictly compound, words. The agglutina- tive are not wholly agglutinative; the Finnish and Turkish classes of the Ural-Altaian family are in cer- tahi instances inflectional, the relational adjunct being fully incorporated with the predicable stem, and having undergone a large amount of attrition for that purpose. Nor, again, are the intiectional languages wholly inflec- tional; Hebrew, for instance, abounds with agglutinative forms, and also avails itself largely of separate particles for the expression of relational ideas ; our own lan- guage, though classed as inflectional, retains nothing more than the vestiges of inflection, and is in many re- spects as isolating and juxtapositional as any language of that class. While, therefore, the classification holds good with regard to the predominant characters of the classes, it does not imply differences of a specific nature. (3.) But, further, the morphological varieties of lan- guage are not confined to the exhibition of the single principle hitherto described. A comparison between the westerly branches of the Ural-Altaian, on the one hand, and the Indo-Kuropcan, on the other, belonging respectively to the agglutinative and inflectional class- es, will show that the quantitative amount of synthesis is fidly as prominent a point of contrast as the quali- tative. The combination of primary and subordinate terms may be more perfect in the Indo-European, but it is more extensively em])loyed in the Ural-Altaian family. The former, for instance, appends to its verbal stems the notions of time, number, person, and occasion- ally of interrogation ; the latter further adds suffixes indicative of negation, hypothesis, causativenesss, re- rtexiveness, and other similar ideas, whereby the word is built up tier on tier to a marvellous extent. The former appends to its substantival stems suffixes of case and number; the latter adds governing particles, ren- dering them post-positional instead of pre-positional, and combining them synthetically with the predicable stem. If, again, we compare the Shemitic with the Indo-European languages, we shall find a morpholog- ical distinction of an ecpially diverse character. In the former the grammatical category is expressed by inter- nal vowel-changes, in the latter by external suffixes. So marked a distinction has not unnaturally been con- stituted the basis of a classification, wherein the lan- guages that adopt this system of internal flection stand by themselves as a separate class, in contradistinction to those which either use terminational additions for the same purpose, or which dispense wholly with inflec- tional forms (Bopp, Compar. Gram, i, 102). The singu- lar use of preformatives in the Coptic language is, again, a morphological peculiarity of a very decided character. Even within the same family, sa\' the Indo-European, each language exhibits an idiosyncrasy in its morpho- logical character whereby it stands out apart from the
other members with a decided impress of individuality. The inference to be drawn from the immber and char- acter of the differences we have noticed is favorable, rather than otherwise, to the theory of an original uni- ty. Starting from the same common ground of mono- syllabic roots, each language-family has carried out its own special line of development, following an original impulse, the causes and nature of which must remain probably forever a matter of conjecture. We can per- ceive, indeed, in a general way, the adaptation of cer- tain forms of speech to certain states of society. The agglutinative languages, for instance, seem to be spe- cially adaiited to the nomadic state by the prominence and distinctness with which they enunciate the leading idea in each word, an arrangement whereby communi- cation would be facilitated between tribes or families that associate oidy at intervals. We might almost im- agine that these languages derived their impress of uniformity and solidity from the monotonous steppes of Central Asia, which have in all ages formed their proper habitat. So, again, the inflectional class reflects cultivated thought and social organization, and its lan- guages have hence been termed "state" or "political." Monosyllabism, on the other hand, is pronounced to be suited to the most primitive stage of thought and so- ciety, wherein the family or the individual is the stand- ard by which things are regidated (^Ililler, Philos. of /list, i, 285). We should hesitate, however, to press this theory as furnishing an adequate explanation of the differences observable in language-families. The Indo-European languages attained their high organiza- tion amid the same scenes and in the same nomad state as those wherein the agglutinative languages were nurt- ured, and we should rather be disposed to regard both the language and the higher social status of the former as the concurrent results of a higher mental organiza- tion.