There was but one people of the ancient world except India which evolved an independent sys tem of grammar. This was Greece. The contrast between the two beginnings is significant of the trend of thought of the two nations. In India the impulse was given by religion, in Greece by philosophy, especially by the Sophists (q.v.), such as Protagoras and Antisthenes. Many of the technical terms still used in grammar were coined by philosophers before formal grammar was known. On the other hand, there is a certain analogue between the two peoples, for the In dians applied grammar first to the Rig-Veda as the Greeks to the text of the Homeric poems. It was therefore the critics of Alexandria and Pergamum who were the first Greek grammarians. Of these the pioneer was Dionysius Thrax, who taught at Rome in the first century B.c., although the Cratylus of Plato, which is of great historic interest in grammar (see ETYMOLOGY), contains some general statements on grammatical topics. Following Plato, Aristotle in the course of his philosophical works distinguished between nouns, verbs, and 'connectives' (Gk. dp6uara, names, sayings; cr6y8eowt, bonds, including pro nouns), and he also recognized the category of case (Gk. irreZo-ts, fall). The Stoic philosophers made important contributions to grammatical knowledge. Thus Posidonius of Rhodes (first century B.c.) gave to the five cases of the Greek noun the names which they still have. The work of Dionysius Thrax, to which allu sion has been made, was, however, the most important from a formal point of view. Sum ming up the work of his predecessors, he set forth a grammar so complete that it formed a model for all succeeding Greek and Latin works on the subject, was the standard moreover for grammars of other languages, as for instance, of an Armenian grammar said by tradition to have been composed in the fifth century, and com mented upon by Yohan Erznkachi in the four teenth century, and still influences in form even the most modern and scientific grammars. While the Sanskrit grammarians, as we have seen, neglected syntax almost entirely, the Greeks paid special attention to this branch of grammar. This is the new contribution in the works of Apollo nius Dyscolus (second century a.n.), one of the greatest grammarians who has ever lived. The grammatical work of the Romans was but an imitation of their Greek models. The two names deserving special mention are those of Varro (B.c. 116-c.28), whose work on the Latin language is of great value, and has some sections on ety mology, the weakest side of Grleco-Roman gram mar, as it is one of the strongest of the Sanskrit, and Priscian (about the close of the fifth cen tury A.D.), whose Grammatical Commentaries, in eighteen books' (the first two on phonology, two more on word-formation, then twelve on inflection, and two on syntax), was the standard authority through the Middle Ages. Medimval times show no progress in grammatical work. It is true that non-classical languages received attention owing to the spread of Christianity, which offered the • Scriptures in the vernaculars of the peoples to whom the new religion was carried. The only grammar in the technical sense of the word, how ever, of this period is the Welsh one, entitled Desparth Edeyrn Ayr, written, as its name im plies, by Edeyrn the Golden-Tongued, in the thir teenth century. With the revival of learning came a change. Latin still exercised its influence over all grammar, and the earliest Renaissance treatises are devoted to Latin in its relation to the vernaculars. One of the first works of this type was the anonymous Tractatus dans modem teutonisandi cases ct tcmpora (Minster, 1451). A single sentence shows its quaint character: Utl cyn verbum neutrunt edder ncutrale heth da rumme also, dot yd noch wereken edder lydent bedudet (and a verbum neutrum or neutrale is thus called, because it expresses neither action nor passivity). The first Greek grammar of this
period was that written by the Byzantine exile Constantine Lascaris (Milan, •476). To give a history of the many classical grammars which appeared from the beginning of the Renaissance to the new epoch inaugurated by the study of Sanskrit would lead too far afield. Suffice it to say that there was no real progress made in scientific grammar of any language until the ex clusive devotion to Latin and Greek gave way at the opening of the nineteenth century to the broader study of other languages as well. The first grammars of the various non-classical lan guages are as follows: Spanish by rElins Auto nius Nebrissensis (1492) ; Arabic by Peter of Alcalri (1505) ; Hebrew by Reuchlin (1506) ; Old Church Slavic, anonymously (1516) ; Ger man by Ickelsamer ( ?) (1527) ; Hungarian by Erdiisi (1539) ; Ethiopic by Victorinus (1548) ; French by Estienne (Stephanus) (1557) ; Syriac by Widinanstadt (1558) ; Polish by Stoienski (Statorins) (1568) ; Bohemian, anonymous (1567) ; Basque, anonymous (1587) ; Tagal by 'Santo Josepho (1610) ; Modern Greek by Por 1 Ms (1638) ; Canarese by Estevano (1640) ; Danish by l'ontoppidan (1648) ; Annamese by Rhodes (1651) ; English by Wallis (1653) ; Lithuanian by Klein (1653) ; Chinese by Kircher (1667) ; Portuguese by Pereira (1072) ; Malay by Raimonds (1674) ; Swedish by Wallen (1682) ; Lettish by Adolphi (1683) ; Gothic, Anglo-Saxon, and Icelandic by lfiekes ( 1689) . The geographical discoveries of this period resulted linguistically in the publication of numerous grammars, in cluding those of Aztec (1555) ; Chiapa (1560) ; Quichua (1560) ; Mixteca (1593) ; Tupt (1595) ; Aymara and Araucanian (1603) ; Huron (1631) ; Guarani (1639) ; Massachusetts (1666) ; Chaynia (1680). The eighteenth century was rich in grammatical contributions. We have grammars of Tamil by Ziegenbalg (1716) ; Hindustani by Schultze (1741) ; Fanti and Accra by Protten (1764) ; Bengali by Mined (1778) ; Marathi, anonymous (1778). The American languages -also received much attention. Thus Moxan was treated by Marbau (1701); Tarsaca by Quixas (1714) ; Koran by Ortego (1732) ; Lulen by Machoni (1732); Totonacan by Bonilla (1752); Eskimo by Egede (1760) ; Otomi by Neve y Molina (1767) ; Abipone by Dobrizhofer (1784) ; Muhegan by Edwards (1788). The latter half of the century was marked by a steady growth of interest in comparative grammar, which was soon to revolutionize the study and the methods of the science. (The history of these branches of linguistics will be treated more properly under PHILOLOGY.) Since the opening of the nineteenth century the history of grammar in its norma tive aspect has been modified to some extent by the comparative method, which is also re sponsible in great part for the historical mode of treatment. The trammels of Latin are less felt, although the simple normative grammar of any Indo-Germanic language is almost of neces sity written along the old familiar scheme. The inadequacy of such a treatment is less apparent, however, in this group of dialects than in those of other language families. Many tongues which have no ease, verb inflection, or the like corre sponding to the Latin must be distorted before they can be forced into the conventional mold. Such incorporating languages as American In dian are cases in point. This difficulty is, it must be confessed, almost unavoidable. The mind of the linguist is, unhappily, so fixed by the forms of his own vernacular and its kindred languages that it is almost a psychic impossibility for him to escape their influence, especially when pre senting the facts of a language in the form of a grammar.