Grammar did not appear as a separate body of knowledge until a late period in the Greek civilization. The merest ground-work of the science had sufficed to meet all the demands of education, of philosophy, and of a literature in course of production; for its development it was necessary to await a period of literary criticism. When the Alexandrian scholars began to compare the idiom of Homer with that of their own day, the requisite stimulus for the scientific study of language was given, and grammar may be regarded as dating from the Alexandrian age.
What was at that time termed grammar, γραμματική, included far more than the modern science; it was the study of literature at large. The grammarian might have nothing to do with what we call grammar, but be a student of textual criticism or mythology. Any sort of study undertaken for the purpose of elucidating the poets was grammatical. Like the modern professor of literature, the only invariable characteristic of the grammarian was his literary point of view.[155]
[155] See Sandys, _History of Classical Scholarship_, pp. 6–10.
The grammatical studies of the Romans were patterned closely after those of the Greeks; the Greek terminology and organization of the science were adopted without change. The Roman interest in the subject was no doubt heightened by the fact that the Roman culture was a bilingual one; thus a broad basis for the study was furnished, and naturally much attention was given to the derivation of words. A large number of scholarly works was produced, and the inferiority of the borrowed Roman culture is perhaps less noticeable in this department than in any other.
It was inevitable that this ‘grammar’, in a condensed form, should come to be used in common education. Its outlines, however, were rather vague, and many of its departments did not lend themselves to the concise statement necessary in a text-book. The first Greek school grammar, the τεχνὴ γραμματικὴ[156] of Dionysius Thrax, which was destined to be the basis of all the school grammars of antiquity, appeared about 80 B.C. It is noticeable that although the definition of grammar that is given[157] is the definition of the grammar of the scholars, the subjects actually treated are little more than the parts of speech. It was natural that there should be this gap between promise and performance. For a long time no doubt this mere outline was filled in by the oral interpretation of the masterpieces in the manner of the scholars; but when these ceased to be studied, in the early medieval period, the study of grammar was confined to the material offered in the text-books.[158]