No CrossRef data available.
Article contents
III Tradition and Legend
Published online by Cambridge University Press: 11 March 2016
Extract
Some of the generic elements we looked at in the last chapter were thematic, some stylistic, and some purely formal. An exhaustive catalogue would be impractical – not to mention tedious. But, taken together, generic elements like these constitute the traditional side of hexameter poetry. Typical features are the ones that define the tradition as a tradition: if poets repeat a feature, then it propagates successfully, survives, and becomes a traditional element of hexameter poetry. Or, to put it more aphoristically: repetition is tradition.
- Type
- Research Article
- Information
- Copyright
- Copyright © The Classical Association 2016
References
1 Lord 1960: 68–98 provides a seminal treatment.
2 More elaborate models are envisaged by Kirk 1962: 95–8 (four phases); Nagy 1996: 29–63 (five phases).
3 For other verbal echoes in Parmenides fr. B 1, see the commentary of Coxon 2009.
4 Empedocles fr. B 133; Choerilus fr. 2 Bernabé (=fr. 316 SH).
5 Kannicht 1982; Lowenstam 1992, 2008: 1–12; Snodgrass 1998: 127–50; Burgess 2001: 35–44, 53–114; Junker 2012: 1–95.
6 For a summary of ancient testimony see Gantz 1993: 445–54 (Gigantomachy), 597–603 (Troilus).
7 Gigantomachy: Cat. fr. 69.89; Pind. Nem. 1.67–9, 7.90, Pyth. 8.12–18; Eur. Ion 205–18 (the temple of Apollo at Delphi), HF 177–80; various later sources, e.g. ps.-Apollod. 1.6.1–2, sch. on Pind. Nem. 1.101, Isthm. 6.47; possibly Theog. 954, Cat. fr. 138.35–6 (=Shield 28–9). Giants (no reference to Gigantomachy): Od. 7.58–9, 206, 10.120; Theog. 50, 185–6. For pre-400 bce pictorial evidence see LIMC, ‘Gigantes’ nos. 1–21, 32–43, 95–388, and (Etruscan) 405–20, 425.
8 E.g. Eumelus, Titanomachy fr. 3 (West's reading Τιτανομαχίαι is an emendation from Γιγαντομαχίαι); Eur. Hec. 466–74 and scholia ad loc., IT 224; sch. Hec. 471 glosses ‘Titans’ as ‘Giants beneath the earth’ (ὑποχθονίων Γιγάντων). See further Gantz 1993: 447.
9 Clay 2003: 113–15.
10 See especially Kossatz-Deissmann, in LIMC s.v. ‘Achilleus’ nos. 206–388 and s.v. ‘Troilos’; LIMC s.v. ‘Achle’ nos. 11–84 (Etruscan).
11 Proclus allots three words to the Troilus episode in his 610-word summary of the Cypria. Unambiguous pictorial treatments of the Troilus episode go back to 700 bce (LIMC s.v. ‘Achilleus’ no. 332a; perhaps ‘Troilos’ no. 7); the date of the Cypria is unknown. In the surviving fragments Janko 1982: 176 sees a stylometric affinity to Hymn. Hom. Aphr. M. L. West 2011b: 232–3, 240 dates it to c.520 bce on linguistic considerations; at 2013: 63–5 he changes his mind and dates ‘older, pre-Cyclic poem(s)’ to the late 600s bce, but the Cyclic epic to c.580–550. Finkelberg 2000 argues for a position broadly consistent with that taken here: that legendary material associated with the Cypria always remained multiform and was never pinned to an authoritative text.
12 Lowenstam 2008: 114–19, discussing possible fourth-century Italian depictions of Lycaon, doubts that even artists of that time were very familiar with the Iliad.
13 Ibyc. frs. S151.40–5, S224 Page; Phrynichus fr. 13 Snell; the fragments of Sophocles’ and Strattis’ Troilus plays.
14 See bibliography in n. 5 above.
15 M. L. West 2013: 130.
16 Danek 1998, esp. 197–201, 251–7.
17 Similarly M. L. West 2005: 40: an Argonautic poem ‘was current only in oral form, or, if it was ever fixed in writing, it disappeared before the Hellenistic age’. We only know of an obscure poem on the building of the Argo ascribed to Epimenides (frs. 57a–59 Bernabé); and Argonautic material used in the Naupactia (frs. 2–10), in Eumelus’ Corinthiaca (frs. 17–23), and in melic poetry (e.g. Pind. Pyth. 4).
18 Alcaeus fr. 308(d); Hymn. Hom. Herm. 514–15. See further Vergados 2012: 553 on line 515.
19 See the new edition of Davies and Finglass 2015.
20 The name ‘hexameter’ implies ‘six feet’: feet are however an abstraction created by later analysis of the metre, and not the historical basis of the metre.
21 On a possible irregularity in this rhythm see pp. 77–8.
22 These divisions are responsible for the main mid-line word break, or caesura: 5 + 7 produces a penthemimeral (‘masculine’) caesura, 5½ + 6½ gives a tritotrochaic (‘feminine’) caesura.
23 On the system's versatility, and authorial control over it (in Homeric epic), see Hainsworth 1968; Friedrich 2007.
24 Hoekstra 1981: 40–53. Note also Hoekstra's earlier (1957) analysis of the Hesiodic formulaic system.
25 Also ἁλιστέϕανον πτολίεθρον (Hymn. Hom. Ap. 410).
26 Also περιτροπέων ἐνιαυτός (Il. 2.295); ἐπιπλομένων ἐνιαυτῶν (Shield 87).
27 This count excludes a supplement at Cat. fr. 69.100.
28 The ἐπήρατον variant also appears at fr. 242.2; Most 2007 rejects that fragment from the Catalogue, but this analysis may add a little weight in favour of its authenticity.
29 On the age of Ζηνός see Janko 1982: 62–3, 87–8.
30 See M. L. West 1966: 79–89 on dialectal elements in the Theogony.
31 Thus Ruijgh 2011, with a catalogue of Arcado-Cypriotisms at 262–7.
32 See M. L. West 1988: 159–65; Haug 2002: 70–106; Tsagalis 2014.
33 Haug 2002: 106, 146, 160.
34 First: though -εω does appear in Homer in places where it is impossible to reconstruct *-ᾱ’ (11× before a consonant, 8× at line-end), more often it appears before a vowel, where *-ᾱ’ is possible (38×). These counts exclude occurrences before trace consonants such as w or s, and bisyllabic -εω on Βορέω/βορέω < *βορέᾱο. Second: unlike the elided form *-ᾱ’, Ionic -εω could in principle be treated as a short syllable by correption (shortening before a vowel in the following word); yet correpted -εω appears only once in extant hexameter, at Hymn. Hom. Aphr. 148: Ἑρμέω ἐμὴ (there Ἑρμέω must be interpreted as a relatively recent adaptation of Ἑρμείω). Third: though the possibility of reconstructing *-ᾱ’ is moderately common in Homeric formulae, in non-Homeric hexameter -εω continues to appear but *-ᾱ’ can never be reconstructed; this suggests that cases of possible *-ᾱ’ in Homer are relatively archaic contexts.
35 For the phase model: Janko 1982: 89–93; Haug 2002: 151–8. Diffusion model: Horrocks 1997: 212–17; Jones 2011; Nikolaev 2013.
36 Haug, in Finkelberg 2011b: i.9, rejects these examples on the grounds that the vowels sit astride the boundary between stem and ending, and this might have delayed the metathesis.
37 As Jones 2011 observes, dual forms of alpha-contracted verbs are regularly treated this way. Alpha-contracted duals do not appear at all in non-Homeric early hexameter.
38 If the ending was originally *-η’ with the omicron elided, the omicron would still have been written: prior to the Hellenistic era elided vowels were normally written explicitly.
39 Reece in Finkelberg 2011b: ii.514–15.
40 Incidentally, this would also destroy all evidence of bisyllabic *-ηο where the omicron was written but elided (see also n. 36 above).
41 The biggest exception is the word ἕως ‘while, until’; but ἕως is anomalous no matter how we look at it. Its spelling implies iambic ⏑ — (1× Homer), yet its actual rhythm is usually trochaic — ⏑ (21× Homer; 1× non-Homeric at Little Il. fr. 32.13 Bernabé); cf. ἕως monosyllabic — (3× Homer), εἵως spondaic — — (23× Homer; 2× non-Homeric at Hymn. Hom. Aphr. 225, Shield 378). *ἧος only appears as a modern emendation. In sixth- and fifth-century Athens, ἕως, εἵως, and *ἧος would all have been written *hεος. One wonders whether ἕως was adopted rather than *ἧος so as to avoid confusion with hος (Athenian for ὅς); both are common at line beginnings. See also Janko 1994: 17–18.
42 Leaf 1902: 391 (Leaf chooses Ἄλτα’).
43 Janko 1994: 385.
44 Iliad: Leaf 1902; Monro and Allen 1920; M. L. West 1998–2000. Odyssey: Allen 1917–19.
45 M. L. West 1978: 61, with a bibliography of other suggestions relating to early orthography.
46 See also Haug, in Finkelberg 2011b: i.117–18, on sporadic vs. systematic Atticisms.
47 Horrocks 2010: 39.
48 M. L. West 1988: 156–9; Horrocks 1997: 201–3; Latacz 2004: 259–67.
49 M. L. West 1973a; Nagy 1974: 49–102; Barnes 1986; Hoekstra 1981: 33–45, esp. 38–9; Ruijgh 2011: 257–8. See also Edwards 1986: 174–88 for a review of older literature on Fränkel's four-colon model.
50 Successful: Andrews 2005: 15–19. Less successful: Mansilla and Bush 2002; Eder 2007; Papakitsos 2011.