1. Linguistics arose in western antiquity from two rather different sources: philosophical debate on the origin and nature of language, and the practical requirements of textual criticism and the teaching of Greek. It generally went under the name of ‘grammar’ (grammatikē), which had at first referred simply to the teaching of literacy, and came later to include what would now be called orthographical phonetics, phonology, morphology, and syntax. Linguistics developed along with other disciplines concerned with language, esp. rhetoric and literary criticism. Several well‐known grammatici engaged in one or both of these other subjects as well.Linguistics began in Greece and was then taken up in the Latin world after the Greek‐speaking countries had fallen under Roman control. There was some independent thought on language in Roman work, esp. with Varro, but in general the Greeks set the pace and the Romans willingly and explicitly followed them. But Latin linguistics did not chronologically just follow Greek linguistics; from around 150 bc and up to the end of the classical era, c. ad 500, when Greek and Latin contacts began to weaken, Greek and Latin scholars were working contemporaneously and often in direct contact with each other.2. Linguistic speculations of a sort are known to have occupied philosophers from the Presocratic period, on such matters as the real‐world correspondences of grammatical tenses and genders. Aristophanes made fun of Socrates engaging in such studies (in Clouds).Two general questions arose: (a), To what extent was language an innate capacity of human beings, and how far was it the result of a tacit convention or social contract? and (b), How far could general statements be made covering large numbers of word‐forms and meanings, and how much individual irregularity must be accepted as inherent in language use? This latter went under the name of analogy and anomaly.Plato's Cratylus deals with the nature of language, and in other dialogues he attributes to Socrates certain linguistic notions such as a nominal‐subject element and a verbal predicate as the basic components of sentences. Aristotle followed Plato's outlines and made reference to linguistic topics in several works.3. It was the Stoics (see stoicism) who recognized linguistics as a separate and essential part of philosophy or dialectic. Breaking away from the Platonic–Aristotelian school, the Stoics favoured the naturalist origin of language, laying stress on the irregularities necessarily found in it. Following their devotion to propositional logic as against the predominantly class‐membership logic of Aristotle, they paid particular attention to syntax.4. Though some details are lacking, we can follow the successive stages in the recognition of different word‐classes (parts of speech) and of the grammatical categories which characterized them. Plato's distinction of nominal subject and verbal predicate was enriched by Aristotle's identification of a complex class of ‘form words’, lacking ostensive meaning and serving to ensure the unity of whole sentences. He also introduced the word ptōsis, ‘falling’, as a technical term for all grammatically relevant word‐form variations. The Stoics later confined the term to its subsequent and current sense of nominal inflexion (Latin cāsus ‘case’). This made possible a further subdivision of the Aristotelian class of form words. Their semantic analysis of Greek verbal tenses into their temporal and aspectual meanings was exploited by Varro in his analysis of Latin tenses.Resulting from the conjoint work of Stoic philosophers and Alexandrian teachers and critics, a system of eight word‐classes—noun, verb, participle, (definite) article, pronoun, preposition, adverb, and conjunction—was established and preserved throughout the Greek grammatical tradition.5. Alexandrian linguistics became the standard model in the Greek and Roman world and very largely formed the basis of both classical and modern grammars of European languages in the Renaissance. Alexandrian grammar, despite its general Aristotelian orientation, was driven less by philosophical considerations than by the needs of teachers of the Greek language and of Greek, esp. Homeric, literature. The Macedonian successor states made themselves responsible for promoting Greek studies in their hitherto non‐Greek territories. This Hellenizing process (see hellenism and hellenization) was taken over and continued by the Romans, and during the four centuries of Roman rule the Greek orientation of education and culture was left intact.Alexandria became a centre of literary and linguistic studies, the latter comprising orthographical phonetics, morphology, syntax, lexical semantics, and dialect studies. Aristarchus 2 was both a grammarian and a Homeric scholar, and his pupil Dionysius Thrax wrote what was probably the first authoritative grammar‐book of the Greek language. It soon became known as ‘The Manual’. A brief textbook entitled Science of Grammar has been attributed to him, but what we have is probably a Byzantine version. In it the sentence and the word were formally defined as the expression of a complete thought and as the minimal unit of syntax, respectively. Next came the eight word‐classes.6. Some Greek grammarians turned their attention to Latin after contacts with the Roman world, and they declared that, with just a few exceptions, the framework of Greek grammar would fit the Latin language. In their recognition of Greek superiority in intellectual matters this was what the Romans wanted to hear, and as far as possible the Alexandrian classes and categories were handed on to the later Latin grammarians.However, Varro was the principal link between Greek and Latin linguistics, a learned man, knowing both languages well, and acquainted with the grammar‐book of Dionysius in its original state. He understood both the Stoic and the Alexandrian views on language and applied them to Latin in his book On the Latin Language. This is not a grammar of Latin, but a lengthy discussion of the language, its structure, vocabulary, and, so far as he could trace it, its history.Varro was the most original thinker about language that we know of in the Latin world. In addition to his application of Stoic semantics to the Latin verb he made an extensive study of word formation and inflexion, drawing on the principle of regularity (‘analogy’), but recognizing existing irregularities as well. In his books he began the process of grouping Latin case forms together, leading to the later establishment of the traditional five declensions. These five were set out by the late Latin grammarians such as Priscian (c.ad 500) several centuries before a comparable simplified account was applied to Greek.Three main differences between Latin and Greek had to be noticed by Varro and others: (a) The Latin ablative case, not found in Greek and recognized by Varro as the ‘sixth’ or ‘Latin’ case. The term ablative was created later by reference to one of its major functions, ‘taking away from’. (b) The absence of a definite article in Latin. (c) The conflation in Latin of the present completive (‘have done’) with the plain past (‘did’), having differential verb forms in Greek but a single form in Latin. This was duly noted by Priscian.7. The first grammarian dealing exclusively with syntax, whose work is, in part, extant, is Apollonius Dyscolus, writing in Alexandria c.ad 200. He was regarded by Priscian as his principal authority, and later Byzantine grammarians in the main wrote summaries and commentaries on the basis of Apollonius' books. The work of these Byzantine Greek grammarians between 500 and 1500 was the main vehicle for the reintroduction of Greek studies in the western Renaissance.