The associative processes and sound changes combine to give origin to inflectional endings. Some are known to have originated in com position; in other cases sounds developing in accordance with purely phonetic laws inde pendent of meaning have later taken on the significance of inflectional elements. Adverbs and conjunctions are commonly developed out of inflectional forms of other parts of speech, especially nouns, adjectives andpronouns, but also, the less commonly, from verbs. The prep osition is only an adverb *in disguise.* Classification of There are a number of principles on which languages may be classified. That most widely known was elaborated by von Humboldt. He distinguished between the outer and inner sides of lan guages, between the movements and the con ceptual forms. He conceived that the latter, being inherent in the human mind, are the same for all nations, but that different peoples expressed them with different degrees of per fection. He considered the most primitive type to be the *isolating* languages, in which the words are all simple roots with nothing resembling inflectional forms. Higher were the *agglutinating* languages which show a partial fusion of roots into loosely united word ele ments, and highest the inflected type. This principle cannot be applied practically to the classification of the languages of the earth, be cause few, if any, of them belong exclusively to any one class. As a matter of fact, both the analytic tendency (toward isolation) and the synthetic (toward inflection) are present in all languages at all times. Now one may pre vail, now the other, as in Old English there were elaborate inflections, while at present English belongs rather to the isolating type, as does the Chinese. In fact languages show such complexity and variety that it may be doubted whether any principle of classification can be consistently applied to them. The best that can at present be done is to put into groups by themselves certain languages which have con spicuous resemblances in vocabulary and ex ternal form. Such clearly defined groups are: The Malayo-Polynesian Group (agglutinative), including Malayan, Melanesian and Polynesian; Bantu or Kafir; Dravidian Group in southern India and Ceylon; Finno-Ugric, comprising Finnic (six languages), Permian, Volga Fin nic and Ugrian, the chief language of which is Magyar; Chinese, an isolating language, as are also the unrelated Anamese, Siamese, Burmese and Tibetan; Turko-Tataric.
There are about 30 groups of languages recognized on the American continent bet their relationships are not all perfectly under stood. The Semitic branch includes, among others, the Assyrian (with Babylonian), He brew, the language of the Old Testament, Phoenician and Arabic (classical and modern). To the Hamitic branch belong Ancient Egyp tian, Coptic, Berber and several languages in Abyssinia and adjacent territory. The Indo European family will be discussed in detail in the next section of this article.
In this article only Indo-European Com parative Philology is treated. The Indo European family of languages contains the fol lowing branches: 1. Aryan or Indo-Iranian group. This com prises (a) the Indian languages, that is, Sanscrit (both Vedic and Classical) with a very rich religious and secular literature; Pracrit, the ancient vernaculars, from which many of the spoken languages of modern India are descended; Pali, the language in which the Buddhistic writings are largely preserved. Modern Gipsy also belongs here. (b) The Persian group, that is, Avestan (the language of the Zend Avesta); Old Persian (the lan guage of the cuneiform inscriptions) ' - Middle Persian or Pehlevi (till about 700 A.D.) ; Mod ern Persian, Kurdic, Ossetian and Baluchi; Parsi, the language of the Fire Worshippers.
2. Armenian, old and modern.
3. Greek, with its many dialects and marvel ous literature. The chief ancient dialects were Attic-Ionic, Doric and IEolic; the main his torical periods are Homeric, Classical Attic, Hellenistic, Byzantine, Modern.
4. The Illyrian group, represented by the Albanian.
5. Italic group, comprising Latin and the Oscan-Umbrian dialects. From spoken Latiu disseminated throughout the Roman Empire the Romance languages have developed.
6. Celtic, including the language of the an cient Gauls, and modern Welsh, Cornish, Scotch and Irish, the last with an important literature.
7. Germanic or Teutonic, represented by (a) old Gothic; (b) Scandinavian (Danish. Swedish, Norwegian and Icelandic) ; (c) West Germanic with its various older dialects from which modern German, Dutch, Flemish and English are sprung.
8. Balto-Slavic includes (a) Lithuanian, Lettish and Old Prussian (now extinct); (b) East-Southern Slavic (Russian, Bulgarian, Serbian, Croatian, Slovenian): Western Slavic (Czechish Bohemian, Wendish, Polish and Polabian).
It is known that all of these had their origin in dialects of one common language, the home of which on linguistic and archeological evi dence is generally conjectured to have been be tween the Baltic and the Caspian seas. Thence they spread by migrations to their later habitats. By the separation thus brought about and in some cases at least by race mixture the differences between the dialects increased until they became distinct languages, which in their turn spread over larger and broke up again into dialects. These later dialects de veloped into still other languages, and so on indefinitely. This process is still going on though more slowly, because of the closer corme munications existing between nations. It is possible, perhaps even likely, that the Celtic and Italic groups arose from one common dia lect, just as did Baltic and Slavic. or Indian and Iranian; in which case we should have to speak of an Italo-Celtic group instead of two separate groups. Each of these Indo-European groups developed separately, no one of theda springing from another. Thus English is just as closely related by origin to Hindi or Bengali or Russian or Czechish as it is to Irish or French, although the extensive borrowing of French words in modern times gives the Eng lish vocabulary a closer resemblance to the French.
The reconstruction of the primitive dialects is accomplished (only, of course, to a very slight degree) by discovering through corn, parson of the related languages what were the original sounds, words, forms and meanings. It must be understood, however, that it is im possible to reconstruct combinations of words. For, while we may reconstruct two different words, we cannot be certain that both were used in the given form in a given dialect at the same time. Furthermore all reconstruc tions are only approximate. But philologists are not distressed on this account, since they are interested not in discovering the starting points of modern forms (which, after all, were only end points as compared with the innumer able forms that for ages preceded them), but in discovering the nature of the changes that took place and the laws by which they were governed.
We know that the early dialects possessed roughly the same group of sounds .as is de scribed in the earlier part of this article. It doubtless possessed at one time a strong stress accent which appears to have brought about such variations in vowel sound as are seen in Greek tithemi, thetos, tethmos, which show the vowel e as long, short and vanished. It also seems that at least in the eastern area this accent was combined with a marked musical or pitch accent. Compound words were freely formed. There was extensive use made of suf fixes and prefixes in derivation. There was very elaborate inflection: three genders (not referring, however, mainly to sex) • three numbers (singular, plural and clual)'; eight case forms were differentiated in the singu lar, six in the plural and four in the dual; the verb had forms distinguishing actions as beginning, ending, progressing, completed or momentary, but gave less accurate expressidn to the time; it had also desideratives, intensives, causatives and iteratives; the verb had also three persons, three numbers, five moods, two voices, active and middle, but apparently no special passive form; there were also various verbal nouns and verbal adjectives. The noun and verb show some striking resemblances in form, which suggest that at a much earlier pe riod they, like the English noun and verb (cf. the word stone used both as a noun and a verb), were only slightly differentiated or not at all.
For the Origin of Language, see SPEECH, GENESIS OF.