1.

Introduction

The ancient Greeks initiated linguistic investigations in several directions. They were the first, and for a long time the only linguistic theoreticians in the world. They also have the credit of being the first Europeans to interest themselves in the accurate study of written texts, and also in fixing the norms concerning the language of their culture. Not only were they the founders of the principles of classical European grammar, but we are also indebted to them for the tradition of grammatical research which was faithfully preserved in later centuries. Their work on linguistic theory developed from their philosophical interest. In order to complete or give more precise form to their philosophical viewpoints they made special efforts to observe the origin of language, the direct relationship between its sound structure and the corresponding meaning, and the possibility of applying logical principles in explaining grammatical forms. All these problems first stated by Greeks are still of interest in presentday linguistics (Ivic, 1965: 2).

2.

The Ancient Greeks

Observation on language, always with reference to the Greek language, is found in records we have of the pre-Socratic philosophers, the fifth century rhetoricians, Socrates, Plato, and Aristotle. Yet, our knowledge of the pre-Socratics and the early
1

rhetoricians is fragmentary and derives from secondary sources. Our knowledge of Socrates is also indirect. He left no writing himself, but arguments and viewpoints are reported in some of the writings of Xenophon and in the more famous Dialogues of Plato, the most important of which is the Cratylus discussing the PhysisNomos controversy to which mention shall be made in the following sections (Robins, 1967: 14). Since the Greek trade and influence of that era were wide spread, there was a practical need to train men of politics and business. Teachers who offered themselves to instruct others in the art of wining victories in debate where considered teachers of the only worthwhile "wisdom". Such were the Sophists (Dinneen, 1967: 7273).

3.

The Sophists

The original meaning of the word "sophist" was "a wise man," since training in rhetoric was equated with wisdom. There was a strong empirical tendency among the Sophists of the fifth century B.C. they attempted to subject everything to measurement, even language study. In their teaching of rhetoric, for example, they recommended the use of rounded sentences, in which phrases and clauses of successive sentences would be of equal length, right down to the last syllable. The Sophists laid the groundwork for the technical vocabulary of rhetoric, much of which, in translation, is still current today. One reason for their success is their empirical
2

approach. They did not merely theorize about what constitute a successful rhetorical composition but observed men in action who were acknowledged masters of art. They analyzed the speeches of the masters in terms of a certain number of units and instructed their students to construct speeches of similar units in similar arrangements. Protagoras is credited with being the first to distinguish sentence types. According to some sources, he listed four prayer, question, statement and command. According to Aristotle, Protagoras was also the first to call attention between gender and tense. Gorgias was one of the first to name and recommend the use of various figures of speech. It is important to note that there is actually no explicit distinction of levels of analysis or different systems in their work. Their main purpose was to teach their students how to convince others and not to discuss "grammar". The Sophist gave us no formal criteria for distinguishing the various units discussed; therefore, their work is not linguistically useful (Ibid: 73).

4.

The Physis-Nomos Controversy

The Greek philosophers debated whether language was governed by "nature" or "convention". This opposition of "nature" and "convention" was a commonplace of Greek philosophical

speculation. To say that a particular institution was natural was to

3

imply that it had its origin in eternal and immutable principles outside man himself (and was therefore inviolable); to say that it was "conventional" implied that it was merely the result of custom and tradition (that is, of some tacit agreement, or "social contract", among the members of the community a "contract" which, since it was made by men, could be broken by men). In the discussion of language, the distinction of "nature" and "convention" was made to turn principally upon the question whether there was any necessary connexion between the meaning of a word and its form (Lyons, 1968: 4) and another question about how words in particular and language in general, acquired meaning. This was discussed din Plato's dialogue Cratylus, where the two positions argued hold that language has meaning arbitrarily and through convention, or that language has some natural connection with the things it is used to discuss. In the dialogue Socrates is asked to settle the dispute between Hermogenes and Cratylus. Cratylus believes that the name of the thing is a consequence of the nature of the thing named, and thus that language has meaning naturally (physei). The phonetic composition of the name, therefore, should mirror the composition of the thing named, so that isolated words can be examined as being true or false in themselves. There should thus be only one correct name for anything. Hermogenes denies this assumption and holds that names stand for things only through convention agreement among the
4

speakers (nomo). Since this agreement is subject to change, any word is the correct word as long as there is agreement about its reference (Dinneen, 1967: 74). The naturalist argument leaned as it must on the weight of onomatopoeia in a vocabulary and on a more general sound symbolism phonological structure of some words, and a good deal of play was made with the fanciful and impossible etymologies of some Greek words by which it was hoped to trace them back to an allegedly "natural" source, since it was admitted on the naturalist side that time had wrought changes in the "first" forms of words. The conventionalists pointed out that vocabulary can be changed at will and that the language is equally efficient once the change has been accepted (Robins, 1967: 18).

5.

Plato

Plato lived in a culture that was fascinated with language and his constant attention to language and discourse must be seen in this context. Speeches were an integral part of many cultural events; poetry was used extensively in education and other fields, while drama provided the focus of civic and religious festivals. Plato disagreed with the Sophists both pedagogically and philosophically, and much of his commentary on language and rhetoric is touched in terms of a dialectical exchange with the Sophists. Plato started with basic questions about the ability of
5

language itself to represent reality. He has Socrates suggest that the most important use of language is an instrument to convey knowledge: "a name is a tool for giving instruction, that is to say, for dividing being" (Cratylus, 388c) (Nienkamp, 1999: 5). He was more concerned about the relation of thought, language and the things talked about than the etymology of individual expressions. He saw that just as some things in nature can go together and others cannot, so too certain words can be correctly combined and others cannot. His attempt to establish a discipline that could deal with such rules was a first attempt to found a formal logic, that is, a system by which we can tell whether combinations are correct or not (Dinneen, 1967: 76). It is Plato who first explicitly distinguished between onomata "nouns" and rhemata "verbs". Yet, these two terms were used differently by Plato from the ones our school-grammars are based on. As defined by Plato, "nouns" were terms that could function in sentences as the subjects of a predication and "verbs" were terms which could express the action or quality predicated. (Roughly speaking, the subject of predication names the things about which something is said, and the predicate is the part of the logos "sentence" which says something about the thing named by the subject). Moreover, what we call now verbs and adjectives were put together in the same class (Lyons, 1968: 10-11). He did not distinguish between the phonetic and phonological levels and seemed to assume that a single sound corresponds to a
6

single letter. In his day the expression grammatike techne, which is sometimes translated "grammar", was not the study of

morphological and syntactic constructions but of letters and their arrangement in words. His study therefor was focused on the lexical system. It is clear that the criteria he used to identify the noun, verb and sentence are not formal, since we cannot apply his terms unless we know both his definitions and meanings of the expressions and in his writings there are many complete utterances in which neither a noun nor a verb appears. He also rejected the hypothesis that things and their names are naturally connected and held that there is only a conventional connection. But he did not investigate the arbitrary, conventional system of language. In Plato's dialogue,

Cratylus, Socrates states that custom is convention or agreement,

and since letters both alike and different can represent things that resemble or differ from each other, it would appear that language is fundamentally conventional. He concludes, "we must admit that both convention and usage contribute to the manifestation of what we have in mind when we speak" (Dinneen, 1967: 76-79).

6.

Aristotle

Aristotle (384-322 B.C.) knew the works of Plato, on which he developed his own thinking. His was probably the most remarkable intellect in antiquity; almost all fields of human knowledge then recognized fell within his scope. Although it is difficult to account
7

for Aristotle's linguistic doctrine with exactitude since we have to assemble it from statements scattered among several works of rhetoric, where it appears incidentally and in other contexts, the outlines of Aristotle's linguistics are fairly clear, and it may be seen that his work marks a development from the position reached by Plato (Robins, 1967: 14-15). Although Aristotle is represented as having objected to most of Plato's work, it is obvious that he accepted the bulk of his master's ideas and then developed them with further insights of his own. Aristotle had a general theory about what there is to know, how men know it, and how they express it in language. He saw several levels on which language can be studied and distinguished the forms of words and of sentences, the meanings of words in isolation and in constructions, and differences between the written and spoken styles of language among other aspects which he included in his De Interpretatione "On Interpretation" (Dinneen, 1976: 79). With regards to grammar, Aristotle kept the Platonic distinction between "nouns" and "verbs", but added to these a further distinct class. These were the "conjunctions"; by this term Aristotle meant all those words which were not members of the major classes, "nouns" and "verbs". He also took over from his predecessors the threefold distinction of gender. He observed that the names of many "things" (the term employed by Protagoras to label the third gender) were grammatically either "masculine" or "feminine" in Greek, and he introduced the terms "intermediate" to refer to the
8

third gender. A more significant advance made by Aristotle was his recognition of the category of tense in the Greek verb: that is to say he noted that certain systematic variations in the forms of the verb could be correlated with such temporal notions as "present" or "past". His teaching on this point, although more explicit than Plato's, is far from clear (Lyons, 1968: 11). He provided definitions of onoma, rhema and logos. Moreover he gave a formal definition of the word as a linguistic unit: a component of the sentence having meaning of its own but not further divisible into meaningful units. He also covered a number of grammatically relevant alterations of a descriptively basic form of a word; oblique cases of nouns, comparative and superlative forms of adjectives, deadjectival adverbs, verbal tenses other than the present and some other verbal inflections (Robins, 1967: 2627). Moreover, he provided descriptions of different kinds of criteria among which are, semantic, formal, phonological, morphological, lexical, grammatical and stylistic (Dinneen, 1967: 86). Robins (1967: 15) states that the Aristotelian age marked the end of an era in Greek history. Aristotle had been appointed tutor to the young Alexander of Macedon, in whose reign Greek administration and Greek ideas spread over the eastern Mediterranean area, and Asia Minor, and a variety of the Attic dialect, or common dialect, became a standard language for government, trade and education over the whole area, gradually displacing the local dialects of earlier periods.
9

7.

The Alexandrians

After his conquests, Alexander founded several colonies and put his favorite lieutenants over them as kings. Two of the most famous were Alexandria in Egypt, governed by Ptolemy, and Pergamon, given to Eumenes. These two cities soon emerged as the centers of Greek culture and were political as well as cultural rivals. Both were famed for their libraries, an invaluable possession in ancient times, and for the universities that grew up around them. Alexandria followed the Aristotelian tradition, and Aristotle's own library was the foundation of the Alexandrian collection. Many scholars settled in Alexandria and conducted research in varied fields (Dinneen, 1967: 94). The city of Alexandria became the center of intense literary and linguistic research. The manuscripts of the authors of the past, in particular those containing the text of the Homeric poems, had by now become intolerably corrupt. The Alexandrians sought to restore the original text and to decide between genuine and spurious works. Admiration for the great literary works of the past encouraged the belief, or rather the misconception, that the language in which they were written was itself inherently "purer", more "correct" than the current colloquial speech of Alexandria and that the "purity" of a language is maintained by the usage of the educated, and "corrupted" by the illiterate. This led to other misconceptions including the tendency to consider the spoken language as dependent on, and derived from the written language.
10

The Alexandrian concern with literature merely reinforced this tendency (Lyons, 1968: 8-9).

8. The Stoics
Among the philosophical schools that grew up in Athens after Aristotle, the most important in the history of linguistics is the Stoic school. The Stoics, founded by Zeno (300 B.C.), worked in a number of fields in which Aristotle had worked, but in certain aspects of philosophy and rhetoric they developed their own methods and doctrines (Robins, 1967: 15). According to Dinneen (1967: 88-94), in their logical work the Stoics were the chief opponents of Aristotle's successors, who were called Peripatetics. In summarizing the Stoics contribution to language we will see that (1) they made a clear advance in the distinctions drawn between the logical and grammatical study of language; (2) there is an increased precision in the technical terminology they used to discuss language; and (3) both of these advances are correlated with the difference between Stoics and Peripatetic logic. Stoics formalized the dichotomy between form and meaning, distinguishing in the language "the signifier" and "the signified", in terms strikingly reminiscent of de Saussure's significant and

signifie'. They also distinguished three meanings of the Greek word gramma "letter": "language consists of 24 letters. But 'letter' can
11

have three meanings, the letter itself, the written sign of the letter and the name of the letter". Moreover they were interested in the study of sound, phone'. They distinguished between the utterances of a language that have no meaning (phonetic "nonsense"), from the utterances of sounds that do have a meaning in the language. One of the factors that impeded the work of the Stoics was their conviction that the tongue is the initiator of sounds, because they were not aware of the role played by the vocal cords. Unlike Aristotle, the Stoics held that there is a natural, necessary connection between the sounds of the language and the things for which the sounds stand, relying heavily on onomatopoeia and sound symbolism, although they could see that the present forms of their language did not justify this claim. Therefore, they undertook to search for the original forms, the roots, or etyma, of current expressions, thus starting the study called etymology.

9.

Anomaly versus Analogy

Robins (1967: 19) states that the opposing views of Aristotle and the Stoics are important since they lead to the second linguistic controversy of antiquity, analogy versus anomaly. According to Dinneen (1967: 94), an important influence on language study came out of a dispute that was not primarily concerned with language. It was based first of all on a difference of opinion about the constitution of the universe and secondarily on how the
12

workings of nature are reflected in language. Some scholars thought that there are no laws or regularities (analogies) to be discovered in nature; others thought that the movements of the stars and the regularities of the seasons are not haphazard (anomalous) but the consequence of inexorable laws. Moreover, Lyons (1968: 6) explains that the controversy between the "naturalists" and the "conventionalists" developed into a dispute as to how far language was "regular". In Greek, as in English, although there are many obvious instances of "regular" patterns in the language, there are also many exceptions. As an example of regular pattern in English, take boy: boys, girl: girls, etc. This is an instance of one kind of "regularity" in language. Greek words for regularity and irregularity in this sense of these terms are "analogy" and "anomaly"; those who maintained that language was essentially systematic and regular are generally called "analogists", and those who took the contrary view are referred to as "anomalists". The anomalists did not deny that there were regularities in the formation of words in language, but pointed to the many instances of irregular words for the formation of which analogical reasoning is of no avail (child: children, etc.). It seems clear that Aristotle favoured analogy and the Stoics favoured anomaly as the dominant theme in language. The division may have been sharpened by the rivalry of Alexandria and Pergamum under Macedonian rule as two main centers of learning, Alexandria dominated by analogists and Pergamum by Stoics (anomalists) (Robins, 1967: 19).
13

The anomaly-analogy controversy lasted for several centuries in the grammatical field, beginning with the work of Xenodotus Philadelphus (284-257 B.C.) and culminating in the work of Apollonius Dyscolus and his son Herodian (A.D. 180). The authoritative codification of the work of the Alexandrians is the grammar of Dionysius Thrax (100 B.C.), which served as a model, both in the sequence of topics and in terminology, for grammars right up to our own day (Dinneen, 1967: 95).

10. Dionysius Thrax

Thrax appears to be the author of the first surviving explicit description of the Greek language. The grammar of Thrax Techn

Grammatik, as it is called, runs to fifteen pages and twenty-five

sections, and comprises a summary account of the structure of Greek. It's only major omission is any statement of Greek syntax, although the word class system and the morphological analysis that are set out in it formed the basis of later syntactic statements. In fact the description given by Thrax was regarded as definitive. It remained a standard work for thirteen centuries, and a modern writer has declared that almost every textbook of English grammar bears evidence of a debt to Thrax. Its orderliness, brevity, and explicitness make it well worth the serious study of anyone with a knowledge of ancient Greek, whether from the point of view of general linguistics or from that of classical scholarship, and a brief

14

notice of its main features is appropriate in any history of linguistics (Robins, 1967: 30-31). In addition to the four Stoic parts of speech Dionysius recognized also the adverb, the participle, the pronoun, and the preposition. All Greek words were classified in terms of case, gender, number,

tense, voice, mood, etc. Dionysius did not deal explicitly with
syntax, the principles according to which words were combined into sentences. This part of the grammatical description of Greek was carried out some three centuries later, less systematically, however, by Apollonius Dyscolus (second century A.D.) (Lyons, 1968: 12).

11. Apollonius Dyscolus

The syntactic gaps in Thrax's grammar were partially filled by the grammar Apollonius Dyscolus who listed some of the features of grammatical agreement among form classes as well as rules for the case government of nouns in construction with prepositions (Dinneen, 1967: 107). Apollonius Dyscolus wrote a large number of books, only a few of which survive, and it would appear that despite the earlier writings on Greek syntax his was the first attempt at a comprehensive theory of syntax systematically applied to the Greek language. His importance together with that was realized by his successors, and the great Latin grammarian, Priscian, some three centuries later referred to him as "the greatest authority on grammar", and explicitly imposed Apollonian
15

methods on his own full-scale description of the Latin language (Robins, 1967: 36).

12. Varro
A contemporary of Dionysius Thrax, Marcus Terentius Varro, called the most learnt man of his time, had been following the quarrel between the analogists of Alexandria and the anomalists of Pergamon. Varro wrote a twenty-five volume work on the Latin language, De Lingua Latina. The views he expressed there about the normal study of language are worth considering, even though they had little influence on his contemporaries, since they show a clear appreciation of some of the differences between form and meaning. The work was written as a consequence of the anomalyanalogy controversy (Dinnen, 1967: 107).

13. Remmius Palaemon and Donatus

Remmius Palaemon was the grammarian who translated Thrax's Grammar into Latin and so set order and terminology of subsequent grammars. One advance over Thrax is Palaemon's restriction that prepositions include only forms that construct with verbs and with the accusative or ablative noun forms. Donatus, a fourth-century grammarian whose short work was the elementary text for schools, also followed Palaemon's organization (and, therefore, Thrax's). One original advance over previous work is his
16

discussion of definite and indefinite pronouns, on the basis of their ability to construct with verbs in all persons or only in the third person (Ibid: 113).

14. Conclusion
The Greek triumph in intellectual civilization is to have done so much in so many fields; their work in logic, ethics, politics, rhetoric, and mathematics, to mention only some subjects, come to mind at once. Their achievement in that part of linguistics, in which they were strongest, namely grammatical theory and grammatical description, is strong enough to deserve and sustain critical examination. It is also such as to inspire our gratitude and admiration (Robins, 1967: 39-40).

17

Bibliography Dinneen, F. (1967). An Introduction to General Linguistics. New York: Holt, Rinehart and Winston, Inc. Ivic, Milka. (1965). Trends in Linguistics. Transl. by Muriel Heppell. The Hague: Mouton. Lyons, J. (1968). Introduction to Theoretical Linguistics. Cambridge: CUP. Nienkamp, J. (1999). Plato on Rhetoric and Language: Four Key

Dialogues. Mahwa: Lawrence Erlbum Associate.

Robins, R. H. (1967). A Short History of Linguistics. London: Longmans.

18