APPROACHING CLASSICAL LANGUAGES:

 CONSIDERATIONS ON METHOD

By Andreas L. Katonis

University of Thessaloniki, Greece

           The Ancient World bequeathed remarkable cultures and, together with them, their unique languages to posterity. Among these, perhaps the most important are the Greek, the Roman and the Indian world, and Greek, Latin and some of India's tongues respectively. Two disciplines are competent to approach the languages of these cultures, as well as any other language of any culture: philology and linguistics.

          No one will deny the major importance of Greek, but in linguistics, which will be our main concern, it is perhaps better not to establish an evaluative scale because every language (like every respective culture) can add some details to the whole, that others may not. Linguistics should not be confused with grammar which is also an important kind of engagement in language. As contrasted with the first, a discipline developed in modern times, grammar has ancient roots both in Greece and in Rome. In India, grammatical tradition seems even to be one of the greatest among the spiritual achievements of that country. But grammar, an effective aid when learning a second language, is prescriptive and is, principally, unable to discover the inherent systemic character of a language as well as its changing disposition. Linguistics, on the contrary, is descriptive and its main task is, exactly, to give explanations.

          This science, however, finds itself under a double pressure: students of other branches will readily admit - as is the personal experience of this writer - that cultivating the field is not an easy job, and yet, on the other hand, there are numerous self-appointed candidates who, at times, vehemently come out in support of ideas launched by themselves or by non-professional authors. Emphasis should be laid on the fact that a good command of a language, and even a native-level mastery, are not identical to the technical knowledge necessary to understand a sophisticated system called human language.

          It is not accidental that linguists try to avoid labels. There are further reasons: think what the result would be if every self-confident view were accepted. Patriotic enthusiasm in language issues, indeed, is frequent. How would European feeling react if it came to hear that some Indian scholars maintain that the primary homeland of every important European language and civilization is to be placed in India, or, that the "roots of every word are Sanskrit"?[1] What is called "traditional views about the origin of Yavanas" is not even known either in Greece or in international scholarship. This ignorance is not welcome but the refutation is correct.[2] The other way round, would Indians and others be pleased, if informed, that in some important European countries the respective languages are thought to be the first and best? Germans think, e.g., that their language has marvelous possibilities in compounding words. This is correct, yet Sanskrit may surpass it. How many of them know this? And how would the whole community feel if they became aware that Hungarian was even considered the unique source for - literally - all the languages of the world, as a "linguist" advocated this cause in the first half of the 19th century? Indeed, this has happened. But competitions of this kind have nothing to do with unbiased scholarship.

          With regard to the remarkable variety of human languages, two points of view are to be taken into consideration: the majority of them can be grouped along a broader or narrower genetic relationship, i.e. languages may have cognates. Apart from linguistic affinity, borrowings among them are also possible. The English vocabulary is quite rich: this is partly due to the borrowings the most important sources for which are perhaps French, Latin, and Greek. To decide which word is a loan-word and which is ancestral, one is supposed to have knowledge of the respective history as well as to have recourse to reference books like detailed monolingual or etymological dictionaries. Between synonyms there are usually differences in register, i.e. language style. An example of this kind is "marriage" (a loan-word from Middle French) and "wedding" (an ancestral Germanic word). A good knowledge of English implies the correct use of such pairs. French, being a Neolatin language in origin, may have handed down, again, pure Latin (learned) elements, as well as any of those which French itself received as borrowings (e.g. Celtic words). Orientation within so complicated relations cannot happen by intuition. One is dependent on the linguistic aids of the languages in question.

          Coincidences or accidental similarities are always possible. In this rich diversity, again, only methodical approach can help. It would be curious if somebody tried to suggest that English and Korean are related just because the word 'man' occurs in both languages with exactly the same meaning (McMahon 1994: 5), or if somebody else would create a lofty theory for a Hungarian case where the equivalents for 'writes', 'balm' and 'Irish' are exactly the same word. In reality, at least two of the three are loan-words.

          To decide if words or any other linguistic element is related to an other element marked by resemblance and consequently, whether a group of the respective languages can be united within broader or narrower sets, a number of verifiable similarities are needed. Systematic similarities, as a rule, refer both to the phonological (i. e. the sound) level and to the morphological one, as well as to other constituents, like the semantic component. It must be underlined that in most cases the history of the respective languages, their literary history and that of the country of their speakers can orientate very usefully. This may be important in the case of languages with a considerable time-depth like Greek e.g.

          A couple of simple examples should elucidate all this. Between English and German, despite the fact that native speakers are unable to communicate without having studied each other's language, there are numerous striking similarities like:

book - Buch, hand - Hand, milk - Milch, son - Sohn, three - drei, six - sechs, mother - Mutter, father - Vater, night - Nacht, and many more. Semantics, too, is important: if book and Buch or three and drei had different meanings, their similarity would be meaningless. But they denote the same things. Some of these words can be analyzed further: indeed, two of the kinship terms have a common element: fa-ther - Va-ter, mo-ther - Mu-tter. This is the formant "-ter", sometimes altered by orthography, found also in English bro-ther and German Bru-der, daugh-ter and Toch-ter, respectively. Orthographic, and slight phonological alterations can be explained. These are shared features and regular and repeated correspondences: similarities too numerous to let us assume that they are borrowings. And if they were, the question should be answered, which language was the borrower and under what conditions. The situation is even clearer with Latin and Greek, both of them being more archaic, a feature more important when dealing with historical data: ma-ter and pa-ter in Latin and ma-ter- (Doric) / me-ter- (Attic) and pa-ter- in Greek. The hyphen after "ter" indicates that the form may undergo changes or may get some other kind of ending (think e.g. of the nominative of both words in Ancient Greek having long, or the vocative having short e).

          More considerations can be added: we already may have been convinced that English and German are related (both of them are called Germanic languages) but Latin and Greek also seem to be related. Like

mother - Mutter, father - Vater etc., we have

Lat. mater - Greek mater/meter, Lat. pater - Greek pater, Lat. frater - Greek phrater/phrator ('brother') etc.

which show both similarities and - to a minor extent - differences. Both of the latter, however, are systematic.

          The words for "mother" begin, in all four languages, with the same sound /m/ whereas the formant "-ter" undergoes slight changes (a feature to be explained, together with the alterations in the vowels, in the fields of English, German and Germanic linguistics respectively). We find both words in Sanskrit: matar- and pitar-, again with some minor differences, the explanation of which is a task of Indo-Iranian and Indic linguistics respectively. The words for "father" begin, both in English and German with /f/ ("v" is /f/ in German) but with /p/ in Latin, Greek, and Indo-Iranian. This, again, is not by chance. There are other examples for the correspondence of the initial /p/ and /f/ like Latin piscis, English fish, German Fisch, meaning all of them the same aquatic animal. There is no place for more examples but the instances given prove one more feature: they demonstrate not only that there is some kind of systematic relation among these languages but also that English and German are closer to each other (exactly for this reason we can comprise them in a family called "Germanic") and more distant from Greek, Latin, and Sanskrit, which, on the other hand, again, are closer to each other than to either of the Germanic languages.

          These examples provide us with the dimension of time: English and German are two important languages used currently, Latin and Greek were used in the same sense two thousand years ago. Both of them survive. Surviving Latin is called Italian, which, though not the same, has undeniable similarities with the first. French, Spanish, Portuguese etc. are also surviving daughter-languages of Latin. The similarities are even more conspicious in the case of Greek: there is no reason to give another name to the language spoken today. It remains 'Greek', though it is useful to speak about Ancient and Modern Greek. If we examine the preceeding stages of English and German (the first can be traced back until about the 8th c. A.D., the second several centuries earlier) we find that systematic similarities are not only existing but are even greater. In other words, they resemble Greek and Latin, and other cognate languages available, more than they do today. Both Greek and Latin, flourishing in Classical Antiquity, again, have their respective past. The case of Latin is somewhat restricted. Greek has a much longer history: the oldest written record extant is a Mycenaean Linear B text dating from the 17th c. B.C. (i.e. about 800 years before Homer!). Again the time factor is important. If we compare earlier dates in both languages we see that similarities are more numerous (and consequently genetic relations are more apparent). Thus, e.g., we find that a postpositive "kwe" ('and') existed, like Classical Latin "-que", in Mycenaean as "-qe" with the same meaning whereas in Classical Greek this was the enclitic "te" ('and'). A Mycenaean example is "a-ko-so-ne-qe" (i.e. "axones-kwe", 'and axes', followed by the numeral 50 indicating the quantity). We may remark that a very considerable number of Mycenaean Linear B words belong to inventories. Latin "que" we find e.g. in the first line of Vergil's Aeneid: "Arma virumque cano" ('Arms I sing and the man'), and "te" is really frequent in Ancient Greek like "andron te theon te" (Iliad, 1, 544) meaning 'of men (=heroes, mortals) and gods'. The alteration between kwe and te appears systematic: cf. e.g. Latin quinque and Greek pente (both meaning '5'). We may add Sanskrit panca and Hindi panc which, again, appear systematic (/c/ is like "ch" in English much). Latin quinque continues in Italian cinque, and Greek pente appears as pende in Modern Greek. If we examined more data from Ancient Italic and Ancient Greek dialects we would find a much richer picture proving the relationship. We may remember now the correspondence between Latin-Greek p and Germanic f: '5' appears as "five" in English and "fünf" in German. These numerals are related though other features, being different, need be explained: Gothic fimf and Old Saxon fif may bring them closer. It is perhaps clear now that where written history stops prehistory begins. But the two notions are relative: where there is Latin linguistic prehistory e.g., still Greek written history is effective. Greek, as a fact, has one of the longest known literary histories.

          Languages have several levels: phonological, morphological, semantic, syntactic. To establish a genetic relationship, correlations at least in the first three of them are needed. Greek, Latin and Sanskrit provide us with many beautiful examples. Take the verb 'carry'. This is 'fero' in Latin, 'phero' in Ancient Greek and 'bharami' in Sanskrit. Stems (roots) can be marked off as follows:

           Two of these have a quite similar past tense (a Praeteritum Imperfectum with the Latin term):

           These forms are "too systematic" to be labelled a borrowing. Borrowings have, much more, an ad hoc nature.

          The case of Latin is somewhat different:

          There are, as is logical, also differences. Latin, though not unrelated, has another ending system explainable within Latin linguistic history, and final -t in Greek (present in Latin and Sanskrit), like in other Greek words as e.g. in leon (genitive: leontos, 'lion'), has dropped. Consequently, the optional -n (the ny ephelkystikon) in Greek e-pher-e(n) is posterior. On the whole, in the examples above we find principled and repeated similarities - features which can be explained. To sum up, we may allude to the so-called Neogrammarian doctrine about the scientific nature of sound-laws or rules (the original word is "Lautgesetz"), put forward in 1878. This means that rules for sound correspondences and changes can be set up which always function under the same circumstances. They must cover all cases whether a few or numerous which fall within their scope to be accounted for. (According to the systemic nature of language, these changes may contribute to subsequent morphological changes). If any data should violate the rule and not be explainable by reference to some linguistic principle the rule is invalidated. Then, one of the possibilities is that we have to deal with an intrusive element. The case of loan-words is exactly of the kind! An example is German Pelz ('fur'). The word is "irregular" because it is a loan-word from Latin pellicius ('[made of] leather'). The word Fell, on the contrary, is regular as is the cognate English fell ('animal skin, hide') showing the f-correspondence. Many textbooks explain this principle like those by Bynon (1977: 22-23) and Trask (1996: 224-228). The first declaration dates back to Osthoff and Brugman (1878: III-XX, mainly XIII-XV). Osthoff and Brugman were professors of Comparative Linguistics and of Sanskrit Philology, the first in Heidelberg, the second in Leipzig.

          Linguistics as a science, exists for more than 200 years now. It was exactly the striking similarities between Indian and European languages, and first of all between Sanskrit, Greek and Latin, that gave rise to linguistic investigations after the famous presidential address of Sir William Jones to the Royal Asiatic Society of Bengal in 1786. Jones found that Sanskrit had a "wonderful structure" which he compared to that of Greek. He has formulated what is thought today to be certain: the "affinity [... may] have sprung from some common source which perhaps no longer exists".[3] Today, the "common source" is labelled by the conventional term "Indo-European". This term for the parental language, not existing any more, was introduced by the English Th. Young in a book review in 1813. Synonymous German "indogermanisch" goes back to J. Klaproth who, after a French wording, used this expression in a book in 1823.

          One of the tasks of research remains to understand why the Indo-Europeans lost awareness of their common ancestry after the expansion (cf. Dumézil 1993: 188), or to investigate if they lost it at all. Haudry thinks that perhaps they did not and cites a passage from Aischylus' Persians which may prove the feeling of affinity (1981: 124). Atossa, the Persian Queen Mother is speaking: "I dreamed that two women in fair vesture, one apparelled in Persian garb, the other in Dorian attire, appeared before mine eyes; both in stature far more striking than are the women of our time, in beauty flawless, sisters of the same race. As for country wherein they dwelt, to one had been assigned by the lot the land of Hellas, to the other that of the barbarians" (181-187)[4]. It is "sisters of the same race" which deserves special attention. The closest linguistic cognates of Greek are exactly the Indo-Iranian languages. Latin may give the impression of being closer but this is due to the fact that both Latin and Greek left a very rich and  important literary heritage having fundamentally influenced European culture and languages and, that Latin, moreover, under Greek cultural influence, borrowed a most considerable Greek vocabulary. Once again the difference between ancestral (core) vocabulary and borrowings must be underlined.

          One should not forget: in 1777, nine years before Jones' address, F.A. Wolf initiated Classical Philology. He enrolled as "Studiosus Philologiae", and not as "Philosophiae", in Göttingen University (see e.g. Sandys 1908, Vol. III, pp. 50-60). He became a renowned Homeric scholar. Classical Philology and Historical Linguistics then, for many decades to come, were considered as "allies". One is inclined to suppose that the two dates are not fortuitous.[5] Together with Classical Philology, Indological Studies were, and still often enough are, considered as constituents of a greater unity: Classical Studies.

          The achievements of both philology and linguistics, reached during the nineteenth and the twentieth centuries, are indeed amazing. Therefore anyone having linguistic ambitions must either thoroughly study introductory works in theorerical, historical and comparative linguistics, special treatises like "Greek" or "Latin Linguistics" included, or must trust himself to specialists who have done this laborious task. Those concerned should be aware of the basic handbooks, and etymological and other dictionaries. In the case of Ancient Greek like those by H. Frisk (GEW) and P. Chantraine (DELG) e.g.[6] , for Latin those by A. Walde - J.B. Hofmann (LEW) and A. Ernout - A. Meillet (DELL), for Indo-Iranian and Sanskrit those by M. Mayrhofer (KEWA and EWAia), for the English language that by C.T. Onions, to mention just one name out of a number, and for German that by Fr. Kluge - E. Seebold and others. The Onions dictionary, e.g., explains what etymology is (see p. V), and, like every other serious dictionary, gives first attestations and datings, e.g. for Latin words (see p. VIII). For Indo-European there is the dictionary by J. Pokorny (IEW), and the latest encyclopedia, abbreviated as EIEC; for Italian that by G. Devoto e.g., for Modern Greek the dictionaries by N. Andriotis, K. Dangitsis, G. Babiniotis, and the dictionary (LKN/ΛKN) by the Institute for Modern Greek Studies in Thessaloniki, to which P. Dorbarakis can be added, then historical grammars, and many more (Babiniotis has two good examples for the latter: 1985 and 1998). The "Epimetro" of the Babiniotis dictionary gives a Greek lexicographical survey from the 16th c. until late 20th c.

          Special dictionaries and handbooks give an inventory and a survey respectively, of what already has been established in research. Every new explanation or suggestion, without reasoned statements and refutations of the former positions, is worthless. The present writer thinks too, that philology and linguistics continue to be "allies". Historical, Comparative, and Indo-European Linguistics are almost unthinkable without (Classical) Philology (a good aid to reliable access to literary sources), whereas traditional Philology only wins when supported by linguistic knowledge.

          As to the reception of the classics, one would readily assume that a similar thinking is only natural in a country like Greece, successor to a very important culture and to a unique language. A recent volume of studies (Christidis et al. 2004) tries to come up with this task. But things are more complicated. Issues of heritage are combined with those of the "language question", another heritage brought by unfavourable historical conditions: how the relations between the classical language and the one used actually should be. This burdens the contributions, as did and still does everyday practice, with a perhaps superflous historical dimension. Such "ankyloses" - to take recourse to the Greek word often heard currently in a figurative sense - have been touched upon by Rhéa Delveroudi on pp. 45 (Greek translation) and 95 (French original) e.g. Similarly, the short paper by Th. Papanguélis (pp. 55-59 and 103-107 respectively), consecrated to the role of Latin in Greece, cannot elude these "ankyloses" either: Latin language and Roman spirit, as seen from Hellas, are a bit different than seen from a third country, as is, similarly, Greece and Greek, on native soil. Actually India, equally heir to a very important linguistic and cultural past, may have similar problems though continuity seems there to be more organic. Discussion of all these problems should be, however, the topic of another contribution.

References