In the search for Indo-European poetry and myth we have to draw on sources of very various character and very various date, from hymns and ritual texts of the second millennium bce to songs and folk-tales recorded in the nineteenth century ce. It might be thought that nothing sound could possibly be built from such diverse materials. But Indo-European linguists are in a similar boat. They work on the one hand with Hittite and Vedic texts that are over three thousand years old, on the other hand with Albanian or Lithuanian, which were first recorded no more than five or six hundred years ago. Yet from data of such unequal antiquity they are able to forge unshakeable structures. The reason is simple. Although languages undergo enormous changes over two or three millennia, and to the casual eye are transformed out of recognition, they may also preserve many highly archaic elements. Even modern English, which cannot compare with Lithuanian as a conservator of ancient morphology, is full of Indo-European vocabulary; it preserves unchanged, almost alone, the original sound [w]; it preserves such old features as ablauting verbs (sing, sang, sung) and free-range preverbs (not easy to get away from). Such things are identifiable as old by surveying the whole system. Similar principles will apply in the present investigation.

As in linguistic reconstruction, we seek to work back to prototypes by comparing data from different branches of the Indo-European tradition. The more widely separated and historically independent the branches, the further back in time their concord should carry us. Within each branch we shall pay greatest attention to the oldest available material, as that is where inherited elements are most likely to appear, and where what seem like significant elements are most likely to be inherited. In some cases, naturally, a genuinely inherited motif may turn up only in a later source, but our emphasis must be on the earlier ones.

The oldest extant texts in Indo-European languages are those in the Anatolian languages Hittite, Luwian, and Palaic, starting in the seventeenth century bce, and written either in cuneiform or in Luwian hieroglyphs.33 The greatest number are in Hittite and date from the New Kingdom, c. 13501200. The majority are prescriptions for rituals; there are also state documents of various kinds, royal annals, prayers, laws, treaties, correspondence,

33 The cuneiform texts are identified by reference to Emmanuel Laroche, Catalogue des textes hittites (Paris 1971) (CTH); the hieroglyphic ones by reference to J. D. Hawkins and others, Corpus of Hieroglyphic Luwian Inscriptions, i-ii (Berlin-New York 1999-2000) (CHLI).

instructions to officials, omen texts, oracles, and a limited number of mythological narratives. The myths, however, mostly seem to be taken over from other, non-Indo-European peoples of the region (Hattics, Hurrians, Babylonians, Canaanites), and have little to offer for the present enterprise. The ritual texts sometimes contain embedded hymns, prayers, or incantations, which may be more relevant. The vocabulary of the languages themselves can throw valuable lights.

Hieroglyphic Luwian inscriptions continue into the first millennium, being most frequent in the first quarter. From later centuries we have inscriptions in other languages of Asia Minor that belong to the Anatolian family, such as Lydian, Lycian, Carian, and Sidetic. They have little to offer in terms of content, but I shall have occasion to mention some of them in the section on metre in Chapter 1.

Of comparable antiquity to the Hittite and Luwian material, and of much richer interest for our purpose, is the Indic. Of prime importance are the 1028 hymns of the Rigveda (RV), thought to have been composed in the Punjab in the period between 1500 and 1000 bce. The collection is arranged in ten books, of which 2-7 are the oldest: these are the so-called Family Books, attributed to poets from half a dozen specific families. Books 1 and 10 are the latest. A second large collection, the Atharvaveda (AV), probably overlaps in time with the later parts of the Rigveda, from which much material is repeated. It is more magical in character, consisting largely of curses, blessings, and charms for various purposes. It exists in two recensions. The better known is that of the Saunaka school, which contains 581 hymns. The other is that of the Paippaladas (AV Paipp.), only parts of which have so far been edited. A later Vedic text that I have occasionally cited is the Black Yajurveda in the recension of the Taittiriyas (Taittiriya Samhita = TS). Later still is the Brhaddevata, an index of the deities of the Rigveda, of interest for the myths that it relates about them.

Apart from this Vedic and para-Vedic literature, India's two great epics, the Mahabharata and the Ramayana, are of some significance. Although later in language and versification than the Vedas and certainly composed at a later time—they grew over a long period, conventionally put between about 400 bce and 400 ce—they clearly continue traditions of narrative poetry going back many centuries.34

34 The subject of the first, the war of the Bharatas, should have taken place (if it was a historical event) around the ninth century bce. For an account of the epics see Puhvel (1987), 68-81 and 89-92, and in general Brockington (1998). He reviews Indo-Europeanists' efforts with the Mahabharata on pp. 67-81.

The oldest documents of Iranian literature are the seventeen hymns of Zarathushtra (Zoroaster) known as the Gathas. They are transmitted as part of the Avesta, the corpus of Parsi sacred books. The three largest components in what has survived of the collection are the Yasna liturgy (Y.), the Yasts or hymns of praise (Yt.), and the Vidêvdût or Vendidûd (Vd.), a book that sets out (in the words of R. C. Zaehner) 'dreary prescriptions concerning ritual purity' and 'impossible punishments for ludicrous crimes'. The Gûthûs form chapters 28-34, 43-51, and 53 of the Yasna. Their language is about as archaic in Indo-Iranian terms as that of the Rigveda, and this persuades some scholars to date them to before 1000 bce; others put them as late as the sixth century. The truth very likely lies between these extremes. Next in age is another section of the Yasna, the 'Gûtha of the Seven Chapters' (Y. 35-41). The remainder of the Avesta, known collectively as the Younger Avesta, dates probably from between the eighth and fourth centuries bce, the Vidêvdût being the latest part.

Zarathushtra lived in eastern Iran, perhaps Drangiana in the south-western part of what is now Afghanistan, and the Avestan language is east Iranian. West Iranian is represented by Old Persian, the language of the royal inscriptions promulgated by Darius I (reigned 521-486) and his successors down to Artaxerxes III (359-338).35

For anything in the nature of Iranian epic we have to wait for the Shah-nama of Firdawsi, a lengthy verse history of the Iranian empire composed about 975-1010. It does not continue a native tradition of epic, but it does embody much ancient myth and folktale.36

From another corner of the Iranian world we have an interesting body of legend recorded at a still later date. This is the heritage of the Ossetes, who form an Indo-European enclave in the ethnic mosaic of the northern Caucasus, speaking a language of the Iranian family. They are believed to be a remnant of the Alans, who were powerful in the region until the eleventh century. Their mythology, which is concerned with a legendary race of heroes called the Narts, was first brought to wider attention by Dumézil, and collections of the material have been published by others more recently.37

35 These are cited from the edition of R. G. Kent, Old Persian. Grammar, Texts, Lexicon (New Haven 1953).

37 Georges Dumézil, Légendes sur les Nartes (Paris 1930); id., Le livre des héros. Légendes sur les Nartes (Paris 1965); id. (1968-73), i. 441-575; id. in Wb. d. Myth., i. 4: Mythologie der kaukasischen und iranischen Volker (1986); Sikojev (1985); Colarusso (2002). See also the highly erudite and informative chapter by H. W. Bailey in A. T. Hatto (ed.), Traditions of Heroic and Epic Poetry (London 1980-9), i. 236-67.

The remaining Indo-European language in which we have texts from the second millennium is Greek. However, these early, non-literary documents in the Linear B script yield little that we can use. A far more rewarding source for Indo-European inheritance is the corpus of Homeric and Hesiodic poetry. Most of it was fixed in writing in the seventh century bce, but some of the material it contains, and the traditional language in which it is expressed, have more ancient roots, reaching back at least into the Mycenaean age. Next in importance are the lyric poets of the period 650-450. It is one of the latest of these (also the most extensively preserved), the Boeotian Pindar, who has the greatest amount of interest to offer. Calvert Watkins has called him 'in many ways the most Indo-European of Greek poets'.

The other languages of the Graeco-Aryan group are Phrygian and Armenian. The Phrygian material is epigraphic, and comes from two separate periods: from the eighth to the fourth century bce (Old Phrygian), and the second to fourth century ce (New Phrygian). Some of the inscriptions are metrical or contain metrical phrases, and certain formulae of elevated diction are recognizable.38 The literary attestation of Armenian begins after the Christianization of the country in the fourth century. There are eight fragments of pre-Christian oral verse on mythological and heroic themes, preserved in quotation, and some other evidence of ancient beliefs.39 Popular oral heroic poetry survived into the twentieth century, and 1939 saw the publication of a national 'folk epic' Sassountsy David, put together from poems transcribed from bardic recitations. It deals with the lives and adventures of four generations of legendary kings who founded and ruled in the city of Sassoun (now Sasun in Turkey). From the formal point of view this Grossepos is an artificial construct, but the materials represent authentic oral tradition.40

For the ancient Thracian and Illyrian peoples the source material is extremely scanty. It consists largely of personal and place names, a few glosses from Classical sources, and one or two inscriptions. To these can be added a larger body of inscriptions from south-east Italy in the Messapic language,

38 Texts are collected by Haas (1966); Claude Brixhe and Michel Lejeune, Corpus des inscriptions paléo-phrygiennes, i—ii (Paris 1984); Vladimir Orel, The Language of Phrygians. Description and Analysis (New York 1997).

39 Collected and translated by L. H. Gray, Revue des Études Arméniennes 6 (1926), 159-67. See also J. R. Russell, ibid. [new series] 20 (1986/7), 253-70; id., Acta Antiqua 32 (1989), 317-30; Ishkol-Kerovpian (1986).

40 Cf. C. M. Bowra, Heroic Poetry (London 1952), 357. I cite the work by page from the translation by A. K. Shalian, David of Sassoun (Athens, Ohio, 1964).

which is generally considered to be Illyrian, and a number of statements by Classical authors about Thracian religion.41

Among the Italic languages Latin naturally takes pride of place. But Latin literature is so pervasively influenced by Greek that it can only be used with the greatest caution as a separate witness to Indo-European tradition. The most promising sources are the earliest poets, who, while by no means innocent of Greek influence, at any rate were the least far removed from older native traditions; religious ritual and language, insofar as they are not based on Greek models; and popular and subliterary material such as charms and incantations. Outside Latin the most notable text is the series of bronze tablets from Iguvium (Gubbio) containing the proceedings of a college of priests, the Atiedian Brethren, with ritual prescriptions and regulations. They date from between 200 and 50 bce, and constitute the principal document of the Umbrian language. Occasional mention will be made of other Italic dialects such as Marrucinian and Venetic.

Celtic evidence, coming as it does from the most westerly of the Indo-European territories, is of especial interest and value as a complement to what can be gathered from Graeco-Aryan sources. There is a gap in time between the continental Celtic material and the insular. The first, consisting of Lepontic, Gaulish, and Celtiberian inscriptions, comes from the Roman period and fades out in the third century ce. The Gaulish inscriptions are the most significant, giving us many names of local deities and sometimes other things of religious interest. Further information on the Celts and their ways is provided by Greek and Latin writers such as Diodorus, Strabo, and Caesar, who were all indebted to the Stoic Posidonius, and some others who were not.42

The insular Celtic material consists mainly (apart from some primitive Irish inscriptions in the Ogam script) of Irish and British literature, and it begins around 600 ce. On the Irish side, besides a not very large quantity of early poems and poetic fragments collected by Kuno Meyer and Enrico Campanile, the main body of pertinent material is narrative prose (with some embedded verse) dealing with heroic and legendary subject matter and dating from the eighth to twelfth centuries. There are four major groups, known as the Ulster and Fenian Cycles, the Cycle of the Kings, and the Mythological

41 See Clemen (1936), 83-92; Detschew (1957); Krahe (1955-64); Mayer (1957-9); Haas (1962); C. Brixhe and A. Panayotou, 'Le thrace', in Françoise Bader (ed.), Langues indo-européennes (Paris 1997), 181-205; C. de Simone and S. Marchesini, Monumenta Linguae Messapicae (2 vols., Wiesbaden 2002).

42 Zwicker (1934-6); Michel Lejeune, Lepontica (Paris 1971); id. and others, Recueil des inscriptions gauloises, i-iv (Paris 1985-2002); Meid (1994); id., Celtiberian Inscriptions (Budapest 1994); Jürgen Untermann, Monumenta Linguarum Hispanicarum, iv: Die tartessischen, keltiberischen und lusitanischen Inschriften (Wiesbaden 1997); Lambert (2003).

Cycle. The most celebrated single work, from the Ulster Cycle, is the Tain bo Cuailnge or Cattle-raid of Cooley (sometimes referred to simply as 'the Tain', though there are also other Cattle-raids). According to Thurneysen it was already known in the first half of the eighth century, but it is preserved in later recensions of the ninth and eleventh.43 Other works to be mentioned by name, all from the twelfth century, are the Dindshenchas, which are prose and verse texts concerned with the lore and legend attached to place names, the Acallam na Senorach (Conversation of the Ancients), a collection of numerous stories and poems with a narrative frame, and the Lebor Gabala Erenn (Book of the Invasions of Ireland), an antiquarian mythical history.

From Britain we have a poetic and prose literature in Early and Middle Welsh—we call it Welsh, but until the Saxons confined it to Wales it was the language of large areas of England and southern Scotland too. The earliest poems are associated with the late sixth-century bards Taliesin and Aneirin. Taliesin was the court poet of Urien, king of Rheged (Cumbria with Dumfries and Kirkcudbright), a champion of the British against the English. Aneirin was a younger contemporary of Taliesin and court poet in Strathclyde. He is credited with Y Gododdin, sometimes misdescribed as a heroic poem, in fact a collection of short praise poems mostly relating to the historic Battle of Catraeth (Catterick).44 The most important prose source is the Mabinogion, a collection of eleven mythical narratives from the tenth and eleventh centuries. Mention will also be made of the Triads of the Isle of Britain, a twelfth-century collection of miscellaneous lore expressed in the form of lists of three.

Finally there is a Gaelic oral tradition represented by songs collected from the Western Isles in the nineteenth century by Alexander Carmichael (Carmina Gadelica, 6 vols., Edinburgh 1928-59). They contain some remarkable survivals of pagan piety.

The earliest evidence relating to the Germanic world comes from Classical authors, most notably from Tacitus' Germania, which drew on a lost work of Pliny the Elder on the German wars. Tacitus mentions the existence of traditional poetry as the Germans' only form of record of the past, and in another work he refers to songs in which the national hero Arminius was commemorated.45 Later, after Christianity brought literacy, we find four separate branches of Germanic poetic tradition: Old High German

43 For surveys of this literature see Thurneysen (1921; on its chronology, 666-70); Dillon (1946), (1948); Koch-Carey (2000).

44 I cite Y Gododdin by the line-numbering of Sir Ifor Williams, Canu Aneirin (1938), which is followed by Koch-Carey (2000), 307-41. The transmitted text is thought to be a mixture of recensions of the seventh and ninth centuries, see Koch-Carey (2000), 304-6.

45 Tac. Germ. 2. 2, Ann. 2. 88; passages from other authors in Clemen (1928).

(Alemannic and Bavarian), Old Saxon, Old English (Anglo-Saxon), and Old Norse. These show mutual similarities of metre and diction that point to a common origin. Insofar as their subject matter is heroic, they look back to the Gothic world of the fourth and fifth centuries; the Gothic lays of that time probably stood in a common line of tradition with those of which Tacitus had written.

Germanic heroic poetry is represented by two stray pages from the Old High German Hildebrandslied, by the likewise eighth-century Beowulf and a few other Old English pieces such as the Finnsburh and Waldere fragments, the Battle of Brunanburh, and the late Battle of Maldon, and by a number of Norse poems of the ninth to twelfth centuries. Most of these last are included in the so-called Elder or Poetic Edda, others are preserved in the prose sagas or other sources, and one, the Biarkamal, in a Latin hexameter version in Saxo Grammaticus' History of the Danes. The ninth- or tenth-century Latin epic Waltharius may also be noticed for its Germanic subject matter, related to that of Waldere.

The Edda further contains mythical poems about the gods and gnomic poetry. Besides the Edda there is a large body of verse by the skalds, professional court poets, composed in a less archaic, highly elaborate style. Two ninth-century skaldic poems of importance for mythology are Thiodolf's Haustlgng and Bragi's Ragnarsdrapa. The Old English corpus includes narrative poems on Christian subjects, reflective and gnomic poems, riddles, and spells. Two spells and a prayer in Old High German, the Merseburg Charms and the Wessobrunn Prayer, will also engage our attention.

Complementing this poetic literature, two substantial prose works of the early thirteenth century are of especial importance. Snorri Sturluson's Prose Edda consists of three parts: Gylfaginning, a delightfully written compendium of Nordic myths about the gods, based on older poetic sources; Skaldska-parmal, a survey of poetic language, with much reference to mythical material; and Hattatal, a treatise on verse-forms. Saxo's History, mentioned above, is a Latin work in sixteen books, of which the first nine in particular are a precious repository of Danish legend.

The Baltic countries were not converted to Christianity till a comparatively recent date—Latvia in the thirteenth century, and Lithuania in the fifteenth— and then only superficially. The consequence is on the one hand that there is no native literature until even later, but on the other hand that pagan gods and mythology remained alive in the popular consciousness long enough to be reported by Christian writers and to leave many traces in songs and ballads. The tradition of these songs is very abundant, especially in Latvia, where over 60,000 were recorded in the nineteenth century. There have also been collections of folk-tales, riddles, and the like. Pagan cult practices were also slow to disappear. Some of them could still be documented in nineteenth-century Latvia.46

The Slavs were converted much earlier, in the ninth and tenth centuries. Again most of the earliest evidence for their native religion and beliefs comes from outsiders' reports. By way of poetic tradition there is an obscure and peculiar Russian epic from 1185-6, the Lay of Igor, besides the oral heroic verse represented by the Russian byliny and the much ampler products of the Serbo-Croat bards. The folklore of the Slavonic peoples is a further source of material.47

Albania has its own poetry and folklore, though it is doubtful how far they can be regarded as representative of a distinct branch of Indo-European tradition, since they are not clearly separated from those of neighbouring lands. The oral epics parallel those of the South Slavs and are in some cases composed by bilingual poets, while the folk-tales often resemble those found in Greece. There remains nevertheless a residue of national mythology.48

0 0

Post a comment