1. As Cervantes, often described as the author of the first novel, has his Canon of Toledo in Don Quijote explain, "La épica tambien puede escrebirse en prosa como en verso" ("The epic, moreover, can be written in prose as well as in verse"). The date of this comment, with which many Renaissance writers would have been in sympathy, is 1605.

2. Bakhtin 1981, 14. For Bakhtin's most definitive statement on epic's refusal of contemporaneity, see pp. 13-14: "In its style, tone and manner of expression, epic discourse is infinitely far removed from discourse of a contemporary about contemporary issues addressed to contemporaries."

3. "Vom Geist der ebraischen Poesie" (1782); English translation available in Simpson 1988.

4. See both "The Storyteller" and, for Benjamin's comments on a work's "aura," "The Work of Art in the Age of Mechanical Reproduction,'' in Benjamin 1969, 83-110; 217-254.

5. Lukacs 1971, chap. 1.

6. See Lord 1960, chap. 2 ("Singers: Performance and Training").

7. See Slyomovics 1987.

8. See Beissinger 1991 Indian oral epic is also performed typically by men who are from the lower strata of society and are even untouchables in many cases; see Blackburn et al. 1989.

9. See Ann L. T. Bergren's suggestive essay (1983), particularly her comment on p. 93 on "blindness and mutilation (with the suggestion of castration) as marks of the male poet" in Greek epic.

10. See Derek Walcott's response to Naipaul's criticism (which he quotes in full) in "The Caribbean: Culture or Mimicry?" (1974), reprinted in Hamner 1993, esp. PP. 52-54.

1. Bakhtin 1981, 8.

2. See Parry 1971; Lord 1960, 1991, 1995.

3. Cf. Nagy 1990,31.

4. Todorov 1990, 20.

5. See Nagy 1990, 362; Flueckiger 1996, 21.

6. Cf. Nagy 1979, 79-93.

7. See Nagy 1990, 9 and 362 n. 127.

8. Lord 1960, 6 (my emphasis).

9. Lord 1991.

10. Reichl 1992.

11. Blackburn et al. 1989.

12. Okpewho 1979.

13. See in general the valuable bibliography of Foley 1985.

14. Martin 1989, 150.

15. Okpewho 1979, 34 (my emphasis).

16. See Edwards and Sienkewicz 1990, 187-189, esp. p. 188: "The epic is considered so important and such a quintessential art form that, if epic performances did not evolve in a particular society, that society was considered to be somehow deficient."

17. Lord 1960, 6.

18. See Okpewho's summary (1979, 240-243).

19. Okpewho 1979, 241, with reference to Finnegan 1970, 109-110. In fairness to Finnegan, I should note that I consider her book, Oral Literature in Africa (1970), a veritable treasure-house of comparative evidence. Okpewho (p. 242) praises Finnegan's book for its reliance on "the relevance of African oral literature for comparative literature in the wide sense" (Finnegan, p. 518).

20. Jensen 1980, 18.

21. Slatkin 1987, 260 (my emphasis); cf. Smith 1974 and Ben-Amos 1976.

22. Absolutist notions of genre can be traced back to Plato: eidos, a word used by Plato in the sense of "genre" (Nagy 1990, 87, 109), is also used in the sense of "form" in his Theory of Forms.

23. Cf. Flueckiger 1996, 132.

24. Martin 1989, 10-26.

25. Ibid., 29 (my emphasis).

26. Martin 1989, 12 (my emphasis).

27. Ibid., 12-42.

28. Cf. Nagy 1996a, 132-133.

29. Martin 1989, 12.

30. Ibid., 26-30.

31. Ibid., 30.

32. Ibid., 30.

33. Nagy 1996a, 121.

34. Nagy 1990, 30.

35. Nagy 1996a, 121.

36. Martin 1989, 30-37.

37. Nagy 1996a, 122.

38. Ibid., 122-128.

39. Ibid., 127.

40. Cf. Nagy 1990, 68 n. 84.

41. Martin 1989, 13. For further details and bibliography, see Nagy 1996a, 128, esp. n. 68.

42. Nagy 1996a, 128.

43. Ibid., 128.

44. Nagy 1990, 388, 390-391; 1996b, 81-82.

45. Nagy 1996b, 110.

46. Nagy 1996a, 37-38.

47. Nagy 1996b, 71.

48. Flueckiger 1989, 40; see also Flueckiger 1996, 131-155, esp. p. 146.

49. Nagy 1996a, 56-57.

50. Flueckiger 1996, 133, summarizing Flueckiger and Sears 1991, 6.

51. Flueckiger 1996, 133, with reference to Bauman 1977, 3.

52. Flueckiger 1996, 133-134). For more on the notion of "episodes," see also Nagy 1996a, 77-82.

53. See Nagy 1997.

1. On the ethnography of commentary and its inseparability from text in performance, see Tedlock 1983; for a tradition of exegesis in Indian epic, see Lutgendorf 1989.

2. Reynolds 1995, 210-211. See also Slyomovics 1987 for an analysis of a single poet's improvisatory adaptations of his text in a constant negotiation over social status.

3. For a concise overview with bibliography, see Whitman 1994. The ancient material is excellently surveyed in the first chapter of Feeney 1991.

4. Porphyry apud schol. B ad Iliad 20.67. My translation is based on the text of Schraeder (1880, 240.14- 241.12, reprinted in chap. 8.2 of Diels and Kranz 1952 [hereafter DK]).

5. Tatian In Graecos 31 (= 8. 1a DK): "Those responsible for the most ancient researches into Homer's poetry, birth, and time are Theagenes of Rhegium, at the time of Cambysses [529-522 B.C.E.], Stesimbrotus the Thasian, Antimachus the Colophonian, Herodotus of Halicarnassus. . . . "

6. Certainly Stoic, but perhaps Neopythagorean too: see Wehrli 1928, 89-91, and Cantarella 1967.

7. Cf. Pfeiffer 1968, 4-5.

8. Most (1994) persuasively argues that when Patroclus upbraids Achilles for his heartlessness by saying: "Your parents were not Thetis and Peleus, but you were born from the sea and the cliffs" ( Iliad 16.33-35), this is an allegoresis of Achilles' parentage—Thetis as sea goddess and Peleus as connected to Mt. Pelion. I thank Philip Hardie for drawing this article to my attention.

9. Feeney 1991, 9. So too Tate 1927, 215 n. 5.

10. That Theagenes wrote about Homer is specified in Porphyry (8.2 DK, quoted above) and the Suda (8.4 DK).

11. Cf. Feeney 1991 8-14, and Svenbro 1984, 101-121.

12. Schol. A ad Iliad 1.381 (= 8.3 DK). For Theagenes as rhapsode, see Wilamowitz-Moellendorff 1932, 219 n. 2. Often this evidence is adduced in support of an ancient interpretation (cf. 8.2 DK) of Theagenes as a protophilologist, downplaying his interest in allegory: see Pfeiffer 1968, 10; Svenbro 1984, 111 Wehrli 1928, 91; Detienne 1962, 65-67. But N. J. Richardson (1975, 65-81, esp. 68) has well shown that grammatical, "philological," allegorical, and other practices coincided in many of these early figures.

13. Such defenses seemed to have flourished in sophistic circles of the fifth century; see Carroll 1895.

14. Tate 1934, 108.

15. Tate 1927, citing B5 DK, on which see Schibli 1990, 100 n. 54.

16. Delatte 1915, 114-115; Wehrli (1928, 90) compares a philosopher often associated with Pythagoreanism: Alcmaeon of nearby Croton (B4 DK). For reservations, see Burkert 1972, 291 n. 67; cautious acceptance: Lamberton 1986, 31-40. See also the following note.

17. Cf. Philolaus (?) 44 B14 DK; first cited in Plato Gorgias 493A and assigned to "a certain wise mythologer from Sicily or Italy," on which see Dodds 1959 ad loc. and pp. 296-299.

18. Indeed, allegoresis and etymology may be seen as two sides of the same coin, as in Burkert's concise formulation: allegory is an etymologized narrative while etymology allegorizes an individual word (1970, 450) Plato plays on such traditions in Phaedrus 252B when he allegorizes passion ( eros ) as "winged" ( pteros ) based on an esoteric hexameter couplet ascribed to Homeric rhapsodes.

19. Nestle 1942, 129-130.

20. Iliad 21.6-7: eera d' Here. Cf. note 9 above.

21. Moralia 19E. The history of terms for allegory is surveyed with bibliography in Whitman 1987, app. I. Cf. Pépin 1958, 87-92.

22. In 60 B.C.E.: Philodemus Rhetoric 1.164.22, 174.24-25, 181.25 (Sudhaus); Cicero Orator 94, De Oratore 3.42, 166-167. Making matters uncertain are passages in Demetrius On Style (99-102, 151, 243), which is dated variously from the third to the first century B.C.E.: see Whitman 1987, 264, and Kennedy 1989, 86, 196.

23. Stesimbrotus comes second to Theagenes in Tatian's catalogue of Homeric allegorizers (see note 5 above); Anaximander is probably the Anaximander the Younger from Miletus who at the end of the fifth century wrote Exegesis of Pythagorean Symbola (58C6 DK = FGH9T1; cf. 59A1. 11 DK); this work appears to have applied to Pythagorean sayings the same kind of allegorical explanation that had been used on Homer: cf. Burkert 1972, 166-175.

24. LSJ s.v. Cf. Pépin 1958, 85-88.

25. On the Derveni papyrus, see now Laks and Most 1997. Pending official publication, I use small Roman numerals for columns of the text printed in Zeitschrift für Papyrologie und Epigraphik 47 (1982) after p. 300. I hope to discuss elsewhere the specific problems of interpretation that the Greek presents.

26. West (1983, 78 n. 14) and Feeney (1991, 22) note in passing that ainittesthai bears the sense "allegorize" here.

27. For the interpreter's error, see West 1983, 85, and, somewhat differently, Rusten 1985, 125.

28. ix. 6-7. For a discussion of this ambiguous sentence, see Rusten 1985, 133-134.

29. v.100-11. Cf. too xiii. 13, where the poet's mention of Zeus as "head" is taken as "expressing something else" ( ainizetai ) probably sovereignty.

30. Discussion in Rusten 1985, 128-130.

31. v.2-4. Cf. the distinction in xix.2 between what is "unclear to the many" but not "to those who rightly understand."

32. iii.5. The word semainein, "to indicate by signs," is used for what the poet means by his obscure expression in xix.7; cf. xxi. 13.

33. iii.4, on which see West 1983, 78 n. 14. In the Republic Plato equates speaking "poetically" with "speaking in riddles" ( einixato, Republic 332B9-C1).

34. Note that Porphyry uses ainittesthai in introducing Pherecydes (B6 DK); cf. Schibli 1990, 99 n. 54 and 117 n. 30, and, more generally, Reinhardt 1960, 35-39, and Whitman 1987, 4.

35. Theognis 681-682: tauta moi einikhtho kekrummena tois' agathois, ginoskoi d' an tis kai kakon an sophis ei. For text and translation, see Nagy 1985, 26ff.; 1990, 149.

36. Frags. 174, 185 in West 1992.

37. Pfeiffer (1968, 5) defines the ainos as a fictional story that has special significance in the present circumstances.

38. Odyssey 14.508. Cf. Nagy 1979, 234-237.

39. Nagy 1985, 22-30; cf. Edmunds 1985, 105-106.

40. Nagy 1979, 222-241, esp. 235-238; 1990, 148 and index s.v. ainos.

41. Aristotle Rhetoric 3.14 (1415b23-24). Cole 1991, 48-49; cf. pp. 55-68.

42. Gentili 1988, 43-44, 197ff., 212-213.

43. Aristotle Rhetoric 2.20 (1393b8-22); cf. 2.21 (1394b34-1395a2), 3-11 (1412a22-26); Nagy 1990, 427. Some apparently meteorological passages from Solon's poetry were, at least later, read as political allegories: he was said to be predicting the tyranny of Peisistratus to the Athenians when he sang: "Just as the force of snow and hail comes from a cloud, thunder comes from bright lightning," so the city should recognize in advance the destruction that threatens when some men become too great. Cf. Diogenes Laertius 1.50 and others cited by West 1992 on Solon frag. 9; cf. frag. 12.

44. Nagy 1990, 149, 192-194, 196-198.

45. See discussion in Cole 1991, 49-54; Cole cites (p. 164 n. 7) Pindar Olympian 2.82ff., 11.10; Nemean 7.12-19; Pythian 2.72 and 3.80ff. (where it is a question of the addressee being wise enough to appreciate the import of Iliad 24.527ff.). Cf. Battisti 1990. I thank Lowell Edmunds for this reference and other suggestions.

46. Alcaeus frag. 6 in Campbell 1982, 239; note too that Theognis's lines on "riddling for the wise" quoted above conclude an allegory of the ship of state.

47. Used similarly in the Derveni papyrus xv.8, where Orpheus "likens" Zeus to Air to express the intelligent and universal ruling principle.

48. Cf. "Longinus" On the Sublime 9.6, 7; Buffière 1956, 105.

49. Cf. Babut 1974, 83-117, esp. 102-103.

50. See further Ford 1997.

51. Simonides frag. 564 in Campbell 1991, 452.

52. Hesiod is sometimes paired with Homer (Xenophanes 21B11 DK) and attacked as an ignorant "teacher of the multitude" ( didaskalos pleiston Hesiodus, Heraclitus 22B57 DK).

53. aeiso xunetoisi (frag. 334 Kern); West 1983, 110n. 82, compares Heraclitus's scorn for the "undiscerning" ( axunetoi, B1; cf. B34 DK) and Pindar's excursus into Orphic eschatology, figured as arrows that "speak to those who are discerning" ( phonaenta sunetoisin, Olympian 2.85).

54. A suggestive analysis of the parables of Jesus in this regard is Kermode 1979. Just before Plato rejects the impious myths of epic, he essays that such tales ought to be buried in silence or at least kept to a "very small audience, bound by pledges of secrecy and requiring extraordinary sacrifices," not the mere sacrifice of a little pig, so as to be heard by "as few people as possible" ( Republic 378A). The rejection of the piglet as the price of initiation is a cutting allusion to the Eleusinian mysteries, which, like Homer, were available to anyone, Greek or foreigner, slave or free.

55. Cf. Plato Protagoras 316D, with the discussion of Richardson 1975, 68-69. The regrettable democratizing implications of their teaching are spelled out at Plato Theaetetus 180D: Socrates ironically contrasts the way the ancients concealed the truth of universal flux from hoi polloi by expressing it (allegorically) through the story that Ocean and Tethys are parents of the gods (cf. Iliad 14.201, 302) with their "wiser" successors who spell out everything in their presentations ( anaphadnon apodeiknumenon ) so that even cobblers can share their wisdom.

56. On Alexandria, see Dawson 1992 for Stoic allegoresis, Long 1992.

57. See the excellent studies of Murrin (1980).

58. So defined in Heraclitus the Rhetor Homeric Questions 5 ( ho gar alla men agoreuon tropos, hetera de hon legei semainon ) Accounts of the rise of Greek allegoresis often begin with this late rhetorical definition: e.g., Buffière 1956; Svenbro 1984, 119-121.

59. E.g., de Man 1979, 1983. On de Man's challenge to the possibility of literary history, see Bush 1991, 35-59.

60. Frye 1957. Useful discussion in Bruns 1992, esp. 83-86.

61. Importantly begun in works like Honig 1959 and Fletcher 1964.

62. Reynolds 1995, 211.

1. For a historical and bibliographical overview of the trickster figure, see Doty and Hynes 1993.

2. Abrahams 1968, 170-178.

3. Lévi-Strauss 1963, 224-226.

4. For additional material on Upper Egyptian epic singers, see Slyomovics 1988, 1987a and b, 1986, and "Praise of God, Praise of Self, Praise of the Islamic People: Arab Epic Narrative in Performance" (forthcoming).

5. 'Awadallah divided the epic into three parts: (1) "the Birth of Abu Zayd" ( milad abu zed ) (2) "the Reconnaisance" ( al-riyada ) and (3) the "Journey Westward" ( tagriba ) For a discussion of the epic's divisions according to oral Egyptian poets, see Reynolds 1995, 16 n. 30.

6. For a translation of the birth sequence according to 'Awadallah, see Slyomovics 1997. The text of 'Awadallah's Handal story is based on my unpublished fieldwork tapes recorded in 1983 in Upper Egypt. A complete set of 'Awadallah's version of Sirat Bani Hilal is available in the Folk Arts Center, Tawfiqiyya, Cairo, Egypt. For text and analysis of a Handal tale collected in a northern Egyptian Nile Delta village by Dwight Reynolds, see Reynolds 1995, 79-87 and 214-215.

7. Abu Zayd has many exploits following the hero patterns described by Lord Raglan, Otto Rank, and Alan Dundes; see Raglan 1934, 1956; Dundes 1980; Rank 1959.

8. Some historians interpret the epic as a literary analogue to the religious and political Islamicization of North Africa: see, for example, Brett 1982.

9. 'Awadallah places Handal's tale in part 1, the "Birth of the Hero." This tale appears in printed versions: the 1948 Cairo "yellow book" edition, Qissat al-Haydabi wa-al 'Uqayli Handal, pp. 181-198.

10. See Culler 1988, 3.

11. Parenthetically, concerning the history of pun-making in Egyptian folklore, so prevalent was the practice of Egyptian punning that a ninth-century Arab rhetorician, al-Safadi, remarked on this propensity: "The poets of Egypt, he writes, excelled in the use of this difficult figure" (the pun, which he called tawriya ) The reason for this, according to al-Safadi, "is that the water of the Nile in Egypt is of an excellent quality so that poets living in these regions are endowed with both delicacy and intellect" (cited in Bonebakker 1966, 74).

12. Barthes 1970.

13. On the ambiguous sexuality of the Upper Egyptian poet, see Slyomovics 1988, chaps. 1 and 2. See also Reynolds 1995, 84.

14. Transliteration protocols for Sa'idi (or Upper Egyptian) Arabic follow Slyomovics 1988, 269-273 ("Appendix A: Notes on Translation and Transliteration").

15. Baker 1978 discusses Tunisian pride in Hilali descent. For a survey of Arab countries where Hilali genealogy is proudly claimed, see Mukhlis 1964, 80-98.

16. For descriptions of punning possibilities, see Redfern 1984 and Brown 1956.

17. For the structural role that Jews and blacks play in 'Awadallah's narrative of the Hilali cycle, see Slyomovics 1988, 60-64.

18. al-Khidr frequently functions as a fairy godmother-like figure; see, for example, Slyomovics 1988, 12.

1. Reference will be made to Serbian, Croatian, Bulgarian, and Romanian literatures.

2. Balkan Christian epic is a term used to distinguish the oral epic sung by Christian singers as opposed to Balkan Muslim singers, the most well-known being those who performed in Bosnia-Hercegovina and who served as many of the informants for Milman Parry and Albert Lord in the earlier twentieth century. In a broader sense, it also refers to the oral epic traditions among the Bulgarians and Romanians. The singing of oral epic, though it continued well into the twentieth century, has gradually been in decline, especially since the Communist period. Epic singers were still performing in Yugoslavia and Bulgaria (to some extent) during the 1960s. In Romania, where the singing of epic is perpetuated by professional traditional Rom (Gypsy) musicians, the genre still continues (though it is definitely dying out). On South Slavic oral epic, see Lord 1960; see also Coote 1978. On Romanian oral epic, see Beissinger 1991.

3. See Young 1976, 28ff.

4. For a discussion of gender issues and Herder, see Fox 1993.

5. See Anderson 1991, 205.

6. See Abrahams 1993, 10, 11.

7. See Koljevic 1991, 5.

8. See Lord 1974, 105. On the influence of Vuk's collections of Serbian oral epic on the developing concepts of nation, see Koljevic 1980.

9. See Karadzic 1969. For an English translation of some of these poems, see Pennington and Levi 1984.

10. Examples of the "jailor's daughter" type of female helper from the Vuk collection include Vuca's wife in "Marko Kraljevic and General Vuca" (who, upon the hero Marko's request, takes "the keys of the dungeon'' and releases three heroes, after which she and Marko negotiate and come to terms). In "Marko Kraljevic and the Daughter of the Arab King," the jailor's daughter "opened the door of the dungeon" and led Marko forth, thereby releasing him. In this and all future references to the Vuk collection, see Karadzic. See also the jailor's daughter figures in "The Story of Bamsi Beyrek of the Grey Horse" from the Turkic epic The Book of Dede Korkutin Lewis 1974.

11. The innkeeper's wife is Janka the alewife in "Marko Kraljevic and Djemo the Mountaineer" (Vuk collection). For the Romanian tradition, see the innkeeper's wife in "Doicin the Sick Man" in Amzulescu 1964.

12. Examples from the Vuk collection include Jerina as a counselor to the hero in "The Wedding of Djuro of Smederevo." Clever maidens also figure in "The Wedding of Marko Kraljevic" and "A Maiden Outwits Marko."

13. Examples of the wise and noble mother abound in the oral epics of the Balkans (in the Vuk collection, for example, see "Marko Kraljevic and Mina of Kostur"). This stereotypical image of the mother fits what might be called the Balkan (or larger Slavic) "cult of the mother."

14. In "The Wedding of King Vukasin" (Vuk collection), for example, a hero's faithful sister (Jevrosima) is juxtaposed to his wife as he offers his sister in marriage to the king.

15. In the Vuk collection, the most striking example is found in "Marko Kraljevic and Mina of Kostur." The Odyssey is, of course, the most obvious and well-known version of this narrative.

16. Vidosava in "The Wedding of King Vukasin" (Vuk collection) is such a traitorous wife. "Banovic Strahinja" also includes a treacherous wife, but she is spared.

17. "Marko Kraljevic and the Arab" (Vuk collection) presents an unconventional woman—the sultan's daughter, who refuses to marry "the Arab," thus challenging her father's orders. Another example is found in Rosanda (who is both beautiful and proud) in ''The Sister of Leka Kapetan." When Rosanda is told to choose a husband from among a group of heroes, she rejects the idea of marriage to any of the men assembled and debunks her brother. Her spirit does not go unpunished; she is cruelly injured.

18. See, for example, the powerful "vila" who prohibits singing in her territory in "Marko Kraljevic and the Vila" (Vuk collection).

19. See "Marko Kraljevic Abolishes the Marriage Tax" (Vuk collection).

20. See Nedic 1965, 362.

21. See Njegos 1986.

22. See Njegos's Srpsko ogledalo ( The Serbian mirror ) first published in 1845. It includes Njegos's own collections of oral poetry in Montenegro, as well as poems written in the "style" of oral epic. For more on the various stages of Njegos's poetic output, see "The Transitional Text" in Lord 1995 and chapter 6 in Lord 1960.

23. One of the themes of the poem concerns the warring factions within the state, which could not unite in the face of the Turkish threat. For a fuller discussion of this, see Koljevic 1991.

24. "Slava" refers to the traditional celebration and feast among Serbian Orthodox families for their patron saint.

25. See Nedic 1965. For a discussion of Vuk and his reflections on history, see also Milosevic-Djordjevic 1994.

26. Njegos wrote two other literary epics in which women play no roles at all: Luca mikrokozma (The ray of the microcosm, 1845) and Lazni Car Scepan Mali (The false czar Stephen the Small, 1851).

27. See Njegos 1986,11. 1913-1919.

28. For a discussion of lament in Montenegro, as well as collected examples, see Dzakovic 1962.

29. See Njegos 1986, 11. 2203-2204.

30. See Turner 1967, 93ff.

31. See Njegos 1986,11. 2216-2217.

32. Speaking of men and social structure in Montenegro, Bette Denich has noted that "property and power are vested exclusively in men" (1974, 244). Furthermore, as for male activity and the external world, "since all public arenas have the potentiality for combat, they are designated as male. The household's external environment is exclusively a male domain" (p. 248).

33. See Njegos 1986, 1. 108.

34. Ibid., 1. 91.

35. Ibid., 1. 1265.

36. Ibid., 11. 1718-1719.

37. See Denich 1974, 246, 251.

38. See Njegos 1986,11. 393-394. Kosovo here refers to the decisive battle in 1389, which heralded the beginning of centuries of Ottoman rule.

39. Njegos 1986, 11. 706-707.

40. Ibid., 11. 979-981. Mihailovich's translation of "vijenac" in the first verse of this excerpt as "garland" obscures its connection to the noun in the title of the epic ("vijenac," otherwise rendered as "wreath").

41. See Sapiro 1993, 40.

42. On the use of rape in the civil war in the former Yugoslavia see Olujic 1995; see also Feldman 1993.

43. See Denich 1974, 250.

44. See Enloe 1989, 62.

45. For an English translation, see Mazuranic 1969a.

46. See Koljevic 1991.

47. See Matejic 1981, 104.

48. See Verdery 1996, 73.

49. Ibid.

50. See Sapiro 1993, 38.

51. See Enloe 1989, 42.

1. For discussion of the wider implications of this tension for the history of epic, see Hardie 1993, chap. 1 ("Closure and Continuation").

2. Ahl 1985, 53; see also pp. 10, 25 (grammatical and physical senses of forma ) 51, 182 ("SIMILitudo, like FORMa, is a grammatical as well as an artistic term in Latin").

3. de Man 1978, 23-24.

4. E.g., Haege 1976, 90-93.

5. Bate 1993, 191; see also p. 236.

6. Barkan 1986, 20-21. See Haege 1976, 85-93, for a systematic treatment of the literalization of simile and metaphor. This was also the subject of a paper presented by James C. Abbott at the 1993 meeting of the American Philological Association.

7. Metamorphoses 10.190-216.

8. A metaphorical approach to the earlier myths of metamorphosis is the fundamental tactic in Forbes Irving's excellent study (199) see, for example, p. 60, on animal metamorphoses: "These stories play upon traditional ideas and metaphors about the relation between men and animals to present an imaginative and mythical expression of familiar human concerns." There is a danger that the accidents of survival may tempt an overly Ovidian exegesis of the earlier material: cf. p. 62: "The imagination Ovid shows in his treatment of the stories can often be a help in interpreting them. He will point to or bring to the surface underlying themes which must have been a part of the appeal of the earlier versions of the story even if they were never made explicit" (my emphasis). Forbes Irving refers to an early essay on the linguistic, metaphorical, and allegorical rationalization of metamorphosis by J. G. L. Mellman ( Commentatio de causis et auctoribus narrationum de mutatis formis [Leipzig, 1786]), which I have not seen.

9. The programmatic function is noted by Solodow 1988, 175-176. The qualifications of Anderson 1989 do not affect the points I wish to make.

10. This happens to be the example used by Black (1962) to illustrate his "interaction view" of metaphor. For a semiproverbial example in Latin, see Plautus Asinaria 495: "A man is a wolf to a man, not a man, when he doesn't know what kind of a person he is."

11. Anderson 1963, 4-5.

12. Feeney 1991, 194-195.

13. Forbes Irving 1990, 10 for an important earlier essay, see Rahn 1953; 1954.

14. For similar reflections, see Feeney 1991, 205 (drawing on Solodow 1988, 162-168).

15. On the plurality of myths of human origin, see Bömer 1969, 70; Feeney 1991, 194.

16. The jingle "monimenta manerent" may play with a Varronian etymology of "memoria a manendo ut manimoria potest esse dicta" ( De Lingua Latina 6.49).

17. See Solodow 1988, 186-188.

18. Schmidt 1991, 60.

19. Giamatti 1968; see pp. 437-443 on the Renaissance image of man as Proteus. Ovid uses Proteus as a figure for the infinitely versatile lover at Ars 1.759-762.

20. Schmidt 1991, 45. Schmidt embraces the paradox at p. 39: "Er ist unveränderlich, nicht trotz, sondern wegen der Verwandlung. Das ist nur scheinbar paradox." It will be clear that I do not agree with Haege's observation in the conclusion to his sharp discussion of the place of man in nature in the Metamorphoses that in Ovid "zwischen dem Wesen des Menschen und dem des Tieres klafft ein Abgrund" (Haege 1976, 191). In a number of cases (Io, Callisto, Actaeon: see Haege, pp. 109115) Ovid does narrate a transformation of the Homeric Circe type, contrasting the continuity of the prior human consciousness within a strange animal body (see Frécaut 1985,115-143), but these cases cannot claim a privileged place within the Ovidian anthropology.

21. Schmidt 1991, 77.

22. The specific meaning in the Homeric context has been much debated (see Heubeck et al. 1988, 69).

23. Murnaghan 1987, 10 n. 12, drawing on the classic studies in Detienne and Vernant 1978. But note that Pucci (1987, 149) attempts to deconstruct the integrity of the Odyssean hero, in terms of a "decentered polytropy," drifting into "panotropy." Polutropos is used only twice in the Odyssey, the second time at 10.330, in Circe's address, where it could be read ironically of Circe's failure to "turn" Odysseus.

24. Servius ad loc.: "Voluere casus, that is to say, to be rolled by misfortunes [casibus uolui]. And it is the figure of hypallage, which occurs whenever words are to be understood by the opposite." For a contrasting combination in Virgil of permanency and "rolling around," cf. Georgics 2.294-295: "immota manet multosque nepotes, / multa uirum uoluens durando saecula uincit" (see also note 55 below).

25. See Hardie 1992, 70, with 77 n. 16. On the relation between bird metamorphoses and the escape prayers of Greek tragedy, see Forbes Irving 1990, 106-107: e.g., Euripides Hippolytus 1290ff.: "Why don't you hide under the ground in shame or change your life for that of a bird to escape this misery?"

26. The Homeric model is Iliad 16.805-806 (Patroclus after Apollo loosens his breastplate): "Confusion seized his mind, his splendid limbs were loosed under him, and he stood there astounded"; Turnus's more complete forgetfulness of (at least his physical) self is achieved through simultaneous imitation of Lucretius 6.1213-1214 (the effects of plague): "And some were even gripped by a total obliviousness, so that they could not even recognize themselves ("atque etiam quosdam capere obliuia rerum / cunctarum, neque se possent cognoscere ut ipsi," which in turn imitates a detail in Thucydides' description of the Athenian plague (2.49.8), a biological catastrophe that entailed the collapse of personal and social identity in the city: "Others on first recovering suffered total loss of memory and did not recognize themselves and their relative"). Note the further ''turns" within Turnus at Aeneid 914-915: "tum pectore sensus /uertuntur uarii."

27. The humor lies partly in the fact that mutata is used of a figurative change within a story of literal transformation.

28. The Io episode displays further traces of the last scenes of the Aeneid: Inachus repeats the lament of Juturna for Turnus (cf. Metamorphoses 1.661-663 with Aeneid 12.879-881); Io's persecution by Juno with an Erinys (1.725) is ended through the reconciliation of Jupiter and Juno, a plotline strongly reminiscent of the last half of the Aeneid; cf. perhaps also the bulls simile of Turnus and Aeneas at Aeneid 12.715-722. Another echo of Turnus's predicament is found in that of Actaeon at 3.198ff.: "fugit Autonoeius ["of the same mind", a speaking matronymic] heros / et se tam celerem cursu miratur in ipso / ut uero uultus et cornua uidit in unda, / 'me miserum!' dicturus erat: uox nulla secuta est.; / ingemuit: uox illa fuit, lacrimaeque per ora / non sua fluxerunt; mens tantum pristina mansit." The Ovidian imitation sensitizes the reader to the connection between the hunted deer simile of Turnus at Aeneid 12.749-755) and the stag of Silvia at 7.483-502, a beast that has undergone a kind of reverse metamorphosis into an almost human creature (and which is alluded to at Metamorphoses 3.240).

29. See Hardie 1993, 34, 40. Note in general that the imperial ideology of the apotheosis of the dead ruler, to which Virgil subscribes as epic poet, also destroys the absolute boundary in the Homeric world picture between human hero and god; on the Ovidian handling of this boundary, see Feeney 1991, 203-205.

30. Solodow 1988, 197-198. Cf. Barkan 1986, 66 (quoted by Feeney 1991, 190): "For all its emphasis upon the blurring of clear categories, metamorphosis is as much concerned with reduction and fixity as with variability and complexity." Similar points had already been made by Albrecht (1961, 179), following Dörrie (1959).

31. Nugent 1985, 30. But Nugent's claim (p. 92) that "the first allegorical work can be seen as a surprisingly sophisticated meditation on some of the conceptual problems which the allegorical genre itself entails" is to overlook the amount of work that has already been done by Ovid and Statius.

32. Cf. Nugent 1985, 54-55, on the deception ( fraus ) of Avaritia: "This transformation has been accomplished explicitly on the linguistic level, by a change in naming [my emphasis]." Cf. also Feeney 1991, 383, on the fraus practised by Statius's Virtus, who also puts on a disguise, in a kind of metamorphosis.

33. This trick with reflexive pronouns is an Ovidian favorite, one of the wordplays that force our attention to the duplicities of personification and other kinds of allegory: see Frécaut 1972, 32, 36.

34. Burrow 1988, 155.

35. Skulsky 1981, 35-36.

36. Feeney 1991, 163.

37. I illustrate the points from Virgil's Allecto and Turnus episode, Ovid's Hunger and Erisychthon episode, and Statius's Virtue and Menoeceus episode. For disguise, see Aeneid 7.415-419; Thebais 10 .639-649; for reflexive pronouns, see Metamorphoses 8.8 19: "seque uiro inspirat"; Thebais 10.673: "seseque in corde reliquit "; for abstract nouns, see Aeneid 7.461-462: "amor ferri, scelerati insania belli, ira"; Metamorphoses 8.828: "ardor edendi'' (semipersonified with furit ); 845: "dira fames" (where dira hints at the Virgilian model for the personification of Fames in the Dira Allecto); Thebais 10.677: "leti . . . amorem." For similes, see Aeneid 7.462-466 (boiling cauldron); Metamorphoses 8.835-839 (insatiable sea and fire, the latter undergoing personification into a voracious "eater"); Thebais 10.674-675 (Menoeceus inspired by Virtue like a cypress struck by lightning; the tree is personified, a process reinforced by allusion to the Ovidian metamorphosis of the boy Cyparissus into the cypress tree: with 10.677 ["letique inuasit amorem"] cf. Metamorphoses 10.132 ["uelle mori statuit"]). Wofford 1992, 136-137, points to the "allegorical force" of the Virgilian Allecto, and to the "figurative density" of the narrative of her action.

38. See Zumwalt 1977.

39. Feeney 1991, 186-187. In Callimachean terms the epic poet is a Jovian thunderer (see note 40).

40. Fama is a kind of chthonic version of Jupiter's thunderbolt, borne by Earth in answer to Jupiter's blasting of her other gigantic children. In line 148, "ira . . .  deorum" can be taken as either an objective genitive (her anger against the gods) or a subjective genitive (the anger of the gods, whose manifestation aroused earth to produce Fama): the grammatical ambiguity mirrors Earth's parodic doubling ofJupiter's activity. Furthermore Aeneid 4.175 alludes to a Lucretian description of the thunderbolt at 6.340-342; and Fama "inflames" at 4.197. Terrorist tactics are common to Jupiter (12.851-2) and Fama (4.187). Ovid comments on the association of the two at Metamorphoses 12.49-52 (the description of Fama): "the mutterings of a low voice, like the noise that comes from the waves of the sea, if you listen at a distance, or like the sound produced by the last rolls of thunder when Jupiter has made the black clouds rattle." This is the literary-historical background to Valerius Flaccus's statement of the hostility of Jupiter to Fama at Argonautica 2.117-122.

41. Fama and Furies: "illam Terra parens ira inritata deorum" ( Aeneid 4.178) hints at the etymology of D ira from dei ira, "anger of god." At Valerius Flaccus Argonautica 2.115ff. Venus's use of Fama is modeled on Juno's use of Allecto: at 128 note Venus's Fury-like power to "uerteredomos," "(over) turn households" (Aeneid 7.336: "odiis uersare domos"), and her adoption of a disguise (like Statius's Virtus) to approach the Lemnian women.

42. Whitman 1987, 53.

43. Whitman (1987) also puts his finger on the way in which Fama comments on a quality of metamorphosis at the heart of the Aeneid: "Unlike Homer's brief figure of social discord, the rhetorical flight of Virgil's "Fama" is a comment on the stylistic and conceptual, as well as emotional, dilemmas of the poem. In a world turning inside out and outside in, it is necessary to find some conceptual category to encompass the action" (p. 53; my emphasis); "Virgil's world is a world in radical transition, always threatening to slip into incomprehensibility"(p. 55).

44. Feeney 1991  187.

45. Hardie 1986, 273-279.

46. Is it accidental that (purely grammatically) there is an ambiguity about the reference of haec in line 434, either to Roma or fama? The epic, annalistic history of Rome is notably one of boundless extent arising out of the most small-scale origins: Fama is thus a peculiarly apt figure for an epic about Rome.

47. See Feeney 1991, 248, on Ovidian Fama: "plasticity of tradition and variable nature of poetic truth." The Theogony is the first allegorical poem of antiquity; could "falsehoods like the truth" be understood of the procedures of personification allegory employed by Hesiod?

48. Svenbro, 1976, 133-134.

49. Virgil's own narrative, however, also licenses a reading of Fama's distortions as merely a selective retelling of the tale; for luxus, cf. Aeneid 1.637: "at domus interior regali splendida luxu"; "turpique cupidine captos" at line 194 is partly the story told at the end of 1.673-674: "quocirca capere ante dolis et cingere flamma / reginam meditor" (Venus addressing her son Cupido).

50. [Heraclitus] Homeric Allegories 79.2: "Epicurus, the Phaeacian philosopher."

51. His speech is itself caught within the fama that is the narrative of the Aeneid (204: dicitur ) as Feeney notes (1991, 187).

52. On the convergence of Jupiter and Fama see note 40 above.

53. And by this act of hearing Jupiter inserts himself in the chain of fama; as mythological actors they operate on the same stage of reality (or unreality): with lines 220-221 ("[Jupiter] turned his eyes to the city of the queen," "oculosque ad moenia torsit / regia") cf. line 196 ("[Fama] turned her course to King Iarbas," ''ad regem cursus detorquet Iarban."

54. For examples of "Gestaltgleichheit" in Ovidian metamorphosis, see Haege 1976, 103-106. At Metamorphoses 4.657ff. Atlas in fact undergoes a more thoroughgoing metamorphosis, as his beard and hair change into (abeunt) woods, etc. Ovid also makes Atlas undergo a change in size (see Haege, pp. 172-173: "Gross-klein"; Acis is the only other example of a change from small to big in the Metamorphoses) analogous to that of Fama: "tum partes altus in omnes / creuit in immensum" (660661).

55. This line alludes to a passage in an earlier work ( Georgics 2.294-295, quoted in note 24 above) describing an oak tree, which cries out to be taken as a figure for the permanence of the Roman state (so Kraggerud 1963, 44 n. 106). In turn "mens immota manet" is the model for Ovid Metamorphoses 2.485: "mens antiqua manet" ("the mind [of the metamorphosed Callisto] remained as it was before"); I readily confess that my ''metamorphic" reading of the Aeneid is not (and cannot now be) innocent of the Ovidian reception of the earlier text. The Ovidian imitation prompts yet further rereadings of both Virgilian passages: at Aeneid 4.449 "mens immota manet" could be understood of the survival of Aeneas's mind despite his figurative metamorphosis (into the tree, into Mount Atlas), and at Georgics 2.294 "immota manet" of the tree becomes questionable in the light of the linguistic violence inflicted on it: of uoluens (literally "rolling them by") R. F. Thomas comments: "The thought is bold, as 'the tree is said to do that which it sees done'" (a figure related to the hypallage in uoluereat Aeneid 1.9; see note 24 above); and on the lines 296-297 Thomas remarks: "The personification is intense, as the tree almost takes on the appearance of an Atlas": Thomas is referring to Aeneid 8.137, but there is an obvious connection to the Atlas of Aeneid 4. A pretty versatile tree!

56. Here I would correct Greene 1970, 83 and 84: "The great shaggy ice-bound figure sustaining the sky is an exemplum of heroic self-denial, of austere exposure to the elements for the sake of the world community. Atlas embodies the qualities which Aeneas has temporarily forgotten"; "This cloak and the idle sword, studded ostentatiously with jasper, point the contrast with Atlas' huge battered head." Greene fails to note the echo of Atlas in the "storm-tossed oak" Aeneas.

1. See Eliot 1957, "What Is a Classic?" particularly pp. 67-70, where he discusses universality and provincialism.

2. Eliot 1957, "Virgil and the Christian World," 127.

3. For comments on local culture, see Eliot 1988, "Notes toward the Definition of Culture," 123-140.

4. Nagy 1979, 7. See also chap. 6, "Lamentation and the Hero," which closes with the distinction between the local specificity of cult-"The hero of cult must be local because it is a fundamental principle in Greek religion that his power is local" and the timeless universality of the Panhellenic tradition and the Homeric poetry that embraces and articulates it.

5. Nagy 1979,, 116.

6. I quote from Timothy Hampton's translation of the passage from the Apologia (Hampton 1990, 93). Hampton's argument that Tasso fears "the ambiguities of particularity" and therefore seeks the "stability and certainty of the universal" has been an important influence on this essay; see Hampton, pp. 88-94.

7. Eliot 1957, 127.

8. All citations and translations from Virgil's Aeneid and Eclogues are taken from Fairclough 1974.

9. Klossowki 1990, 128.

10. That Virgil's sympathies may well lie with Meliboeus in the dialogue is indicated by a line relevant to the shepherds' participation in local cult. At one point, Meliboeus explicitly calls attention to the fact that the local fountains Tityrus rejected do, in fact, hold Tityrus dear; when Tityrus was "absent" in Rome, Meliboeus claims, the very trees and sacred fountains called for him ("Tityrus hinc aberat. ipsae te, Tityre, pinus, / ipsi te fontes, ipsa haec arbusta vocabant," 38-39).

11. "et iam summa procul villarum culmina fumant/ maioresque cadunt altis de montibus umbrae" ("Even now the house-tops yonder are smoking and longer shadows fall from the mountain-heights," 82-83).

12. I owe this insight to Laura McClure, of the Classics Department at the University of Wisconsin.

13. "iam tum religio pavidos terrebat agrestis/ dira loci, iam tum silvam saxumque tremebant" (8.349-350). Might we not note in the reference to the pavidos agrestis and the dual emphasis on their terror some condescension on Virgil's part when speaking of those rustics who allow themselves to become so frightened by a mysterious landscape?

14. Book 8 is more ironic than I have indicated; the supposedly local religion of the Capitoline Hill turns out, through the vehicle of the Arcadians themselves, to have been a Greek importation.

15. Perhaps one of the most telling lines in Aeneid 4, when Aeneas is building a city in Carthage, is that which for many critics makes Dido a dissembling woman desperate for marriage with Aeneas. After the two have begun their affair, Dido refers to their relationship as a "marriage" ("coniugium vocat; hoc praetexit nomine culpam," "She calls it marriage; with that word she covers her fault," 4.172). But in a city such as Carthage, which worships Juno as its preeminent deity, the ritual that Juno effects around the cave in which Aeneas and Dido first make love enacts for Dido the equivalent of a marriage rite. For Aeneas, it does not. In this clash, one culture is necessarily silenced (''Urbs antiqua fuit. . . Karthago"—"There was an ancient city called Carthage," we learn at the opening of the Aeneid [1.11-12]), and the other necessarily privileged.

16. For suggestive interpretations of the Italian deification of hybrid figures who supposedly led their people from savagery to civilization, see Brelich 1976.

17. See Williams 1989, 215.

18. More specifically, Freud saw in Tasso's version of Virgil's text a manifestation of the universal compulsion to repeat. See Freud 1961, 16, where Freud suggests that Tancredi's wounding of Clorinda for a second time in the enchanted forest is "the most moving poetic picture" of "the compulsion to repeat." For two recent readings of the poem that draw on Freud's insights, see Ferguson 1982 and Bellamy 1994.

19. See Galinsky 1969, 141-190, and Momigliano 1987, 272-274.

20. See Susanne Wofford's observation that Virgil uses "Roman place names as tropes to legitimize Aeneas's claims to the Italian land retrospectively" (1992, 180). Wofford also calls attention to the actual distance of Aeneas and the Trojans from the land; and it is the sacrifice of Palinurus that "indicates the cost not only of creating such an aetiology for the land . . . but also of a certain type of poetic figuration itself-the cost to poet and hero of possessing the land as a poetic or a political conquest" (p. 181).

21. Galinsky 1969, 190.

22. See Hampton 1990, 89: Hampton quotes from the Apologia ("The poet will have brought the truth and particularity of history to verisimilitude and universality, which is proper to his art") and comments: "As epic poetry becomes poetry, 'considering' things in their universality, historical particularity loses its significance. As events and characters are placed into the unity of a plot they lose their essentially historical character and become poetic—that is, in the parlance of both Tasso and Aristotle, philosophical." Quint 1983 also discusses Tasso's avoidance of historical particularity in the Gerusalemme liberata, although in the context of Tasso's reliance on Platonism rather than on Aristotle. Quint 1993 is a more persuasive reading of Tasso's construction of a "universal" epic, as Quint elaborates Tasso's only partially successful suppression of the sectarian and nationalistic boundaries that divided late sixteenth-century Europe.

23. Tasso 1982 (the edition of Fredi Chiappelli). English translations of the Liberata are taken from Tasso 1987 (Ralph Nash's translation); I have made minor revisions throughout.

24. There are admittedly several other moments in the Liberata when it is Christian blood that "flows" in a manner reminiscent of Christ's. When the warrior Sveno is killed in the desert en route to Jerusalem, his blood becomes a river ("di sangue un rio," 8. 19). Somewhat more problematically, in a passage I will address below, when Clorinda is wounded by Tancredi—a Clorinda who is still, technically, a Muslim, although within several lines she will be baptized by Tancredi himself—her blood is also said to flow like a river.

25. The Conquistata will make this paradox less likely-—the poem ends not with bleeding bodies, but with a lengthy procession and pomp that rivals the "ornamenti barbarici e pompe" stained by the pagans' flowing blood. In fact, in the Conquistata, nature itself crowns Goffredo's victory as he enters the holy city: "E' giá tranquillo il mar, sereno il vento,/ l'aria piú chiara assai ch'ella non suole;/ tanto col vincitore il ciel s'allegra,/ e la natura, dianzi afflitta ed egra" ("The sea is calm, the wind serene, the air more pure than it is wont to be; thus sky and nature alike, once so ill and afflicted, rejoice with the conqueror," 24.132). The fugitives on whom the penultimate stanza of the Liberata had dwelled are relegated to a verse tucked into the middle of the last canto. In lines so altered as to be unrecognizable, they are compared to "belve in fèro ludo/ cinte d'intorno, o 'n sanguinosa caccia" (''wild animals surrounded in fierce struggle, or in bloody hunt," 24.118). Any equation that might have been made between Christ's bleeding body and the fuggitivi can no longer be imagined. Citations are from Tasso 1934 (Bonfigli's edition of the Gerusalemme conquistata ) translations are my own.

26. The "Allegoria del Poema" can be found in Tasso 1875, 1: 301 (Guasti's edition of Tasso's prose works); the translation is from Tasso 1987, 469 (Nash's translation). As Tasso scholars have noted, the "Allegoria" has a singularly complicated relationship to the poem itself; Tasso's letters from the period suggest that it was largely a creation to appease the Roman inquisitor, Silvio Antoniani, who needed to approve the text before it could be published. See, among others, Murrin 1980, 121-128, and Derla 1978. Rhu 1993 provides a translation of the "Allegoria," as well as a splendid introduction to the young Tasso's theoretical works.

27. The phrase is from the later Conquistata, 24.118.

28. Berger 1988 offers a lively discussion of an animism that Protestants such as Spenser associated with Catholicism; see p. 78 in particular: "The Catholic abuses suggested in the early cantos [of Book I of The Faerie Queene] are traced back and reduced to an archaic pagan sensibility which projects anthropomorphic idols as if it never received the Word. The historical failure represented as the Church of the Middle Ages is seen as a betrayal of the original Gospel experience and as a regression to the inherent tendency of the silva vanus, the Flesh, to resist enlightenment." Berger's comments are particularly suggestive as regards the enchanted forest in the Liberata. More generally, see Eire 1986.

29. On the effects of Tridentine doctrine on popular culture in Italy, see Niccoli 1987.

30. There is an interesting parallel with the Inferno, in which Dante's miserable suicide, Piero della Vigna, parodies his own would-be crucifixion: he is trapped like Tasso's Clorinda in a "body or tomb, I know not which to say" (13.43). But Tasso's episode of the bleeding trunk is complicated in ways that Dante's is not. As we know from the formidable announcement posted at the gate of hell, Inferno is God's handiwork, and the fact that hell's inhabitants parody the holiest mysteries of Christianity does not subvert those mysteries but attests to their universality: even in hell, even among pagans, one is forced to acknowledge the truth of Christ's birth, death, and resurrection, and Dante's sinners do so without realizing to what they are attesting. The forest within which Clorinda speaks is the result of horrible charms, "too awful to say," of the Muslim sorcerer who was born a Christian, and unlike the pilgrim Dante, Tancredi has no Virgilian guide to assist him.

31. Barberi-Squarotti 1993, 249, observes rightly that Tancredi can't convince himself that Clorinda is simply a diabolical apparition: "è, anzí, una verità che va oltre il fatto e la vicenda.". The observation is an important one, for it suggests that Tancredi insists on seeing the tree as a material realm inhabited by a transcendent being and thus as a vehicle for incarnation.

32. See the third selection from I Miracoli della Vergine, entitled "D'uno che vendé ciò che egli aveva e dièllo a' poveri" (Of one that sold all he had and gave it to the poor), in de Luca 1977, 4: 730-732.

33. Ugone tells Goffredo: "If high Providence elected you as the chief captain of the venture, he also destined that [Rinaldo] must be the sovereign executor of your commands" (14.13).

34. From the very start of the poem, Tasso depicts Clorinda as adverse to Ismeno's patently un-Islamic practices. In canto 2, the sorcerer is condemned not by the narrator for his confounding of "due leggi," but by Clorinda, when she chastises Jerusalem's king, Aladino, for letting Ismeno persuade him to remove an image of Mary from the Christians' temple and place it in the mosque instead, "so that it will be a fated protection for these gates" (2.6). Clorinda angrily insists that the wizard has little reverence for Muslim law ("Fu de le nostre leggi irriverenza/ quell'opra far che persuase il mago"), and she blames Ismeno for trying to "contaminate'' Islam "con nova / religion" (2.50-51) that believes in the power and efficacy of images. Such an episode shows that Tasso was well aware of some of the central practices of Islam; it also suggests that the demonization of Islam in the text proceeds not from ignorance but from ideology.

35. My thanks to Paul Bucklin, whose senior thesis on Tasso suggestively discusses ways in which the Gerusalemme liberata makes Tasso's predecessors, Dante included, suspicious heretics in light of the "true" Counter-Reformation faith.

1. A version of this essay appears in Flueckiger 1996, where the Candaini epic is contextualized within the broader regional repertoire of Chhattisgarh; it is a substantially revised version of an essay that appears in Oral Epics in India (Blackburn et. al, 1989).

2. See Chapter 1, "Epic as Genre."

3. See, for example, Oinas 1972; Chadwick and Zhirmunsky 1969; Johnson 1980.

4. Blackburn and Flueckiger 1989, 6.

5. Flueckiger and Sears 1991 6; Bauman 1977, 3.

6. Blackburn and Flueckiger 1989, 11.

7. The Awadhi variant was recorded in Allahabad District, U.P., and published as The Hindi Oral Epic Loriki (1979); the Bhojpuri variant was recorded in Benaras, U.P., and published as The Hindi Oral Epic Canaini (1982).

8. Elwin 1946, 338-370; episodes of Elwin's version are surprisingly similar to the episodes I heard in performance, even though they were documented forty years earlier.

9. Miller 1981, 71; Flueckiger 1996, 8-9.

10. Pandey 1979, 17.

11. Coccari 1984.

12. Mandelbaum 1972, 444.

13. One such volume is Khedkar 1959.

14. See Flueckiger 1996 for an expanded discussion of the epic episodes that illustrate the martial and caste emphases of the U.P. variant.

15. I add the characterization "male," since women are not part of its primary audiences and may listen to its performance only when it is held in a setting that allows them to "overhear" from behind a curtain or wall. According to Pandey (personal communication), Ahir women may know the general outline of the narrative but have not incorporated its characters and plot into their own female performance genres.

16. Pandey, oral communication, June 1982. I heard a similar saying in Chhattisgarh regarding the Mahabharata (except argument was substituted for battle) in an explanation for why Ramayana performances were more common-a wonderful indigenous articulation of the creative power of performance.

17. Many oral epics in India are published in these bazaar pamphlet forms; the Chhattisgarhi Candaini, however, has not yet been so published. This particular U.P. publication and its cover illustration seem to be patterned after the popular pamphlets of another martial epic performed in U.P., the Alha Kand, whose episodes are also named after its numerous battles.

18. See Babb 1975, 36-37, for a description of mataras celebrated in the Chhattisgarhi plains.

19. This image is changing, however. During my last trip to Chhattisgarh in the summer of 1993, I heard many complaints from village landlords that Rauts were no longer willing to "serve" the village, that they were choosing to commute to the city for work instead. It has left many landowners desperate for "servants" ( naukar ), and many are being imported from the neighboring province of Orissa, where there is high unemployment and hence a willingness to relocate for work. It is doubtful that the Rauts will reappropriate the epic now to promote this newly emerging identity, since they have already abandoned the epic as performers, although they still participate as audience members of the Chhattisgarhi regional folklore community.

20. Rauts do have other narrative performance traditions whose musical accompaniment—five-to six-foot bamboo flutes—give the genre its name, bas git; and bas git remains specifically associated with the caste, although it is listed in the core Chhattisgarhi folk repertoire referred to earlier.

21. One informant told me that the primary difference between the Candaini and Pandvani epic traditions was the characteristic line ending of mor and tor of Candaini and bhaiya or bhaige (literally, "brother") of Pandvani.

22. Many naca performers are able, therefore, to perform in the git style and may do so for their own entertainment.

23. See Flueckiger 1988 for a description of one performer who experimented with combining elements of git and naca in a public performance for which there were not sufficient funds to hire an entire naca troupe.

24. See Narayana Rao (forthcoming) for a discussion of the Sanskrit classification of low castes and women within a single category.

25. Elwin 1946, 349.

26. Loose hair in Chhattisgarh has sexual connotations unless framed in ritual contexts of mourning or goddess possession.

27. Elwin 1946, 345.

28. Ibid.

29. Pandey, personal communication, June 1982.

30. There are genres other than epic that have been appropriated by folklore groups and communities within the region, however, in a self-conscious way. For example, in 1985 one village headman talked specifically about the role he thought folk festivals could play in establishing a sense of village identity and improving morale. He told me he had introduced the festival of gaura to his village several years ago, one of the performance genres consistently mentioned in the repertoire list of "what we in Chhattisgarh celebrate." He had asked several daughters-in-law who came from villages in which gaura was traditionally celebrated if they would be willing to introduce the festival to their village of marriage. He thought this particular festival would be appropriate because of the numerous folklore groups it could involve and the large public procession to the village tank that ends the festival. After several years of its performance in his village, the headman seemed satisfied with its results.

31. Bose and Bhattacharjee 1984.

32. Flueckiger 1988.

1. See, for example, Briggs 1988, Bauman 1977, and Hymes 1971. For an earlier and broader statement, see Goffman 1974, particularly chap. 10, "Breaking Frame."

2. See Hymes 1981, "Breakthrough into Performance" and "Breakthrough into Performance Revisited."

3. For an overview of these materials, see Reynolds 1989.

4. See Bird 1972; Lord 1974; Reichl 1992.

5. The authoritative role of the first-person utterance has been an enduring feature of Arabic literature from the earliest periods onward, particularly in genres that are close to, or imitate, oral traditions. See Reynolds 1991.

6. Reynolds 1995, chap. 5, "The Interplay of Genres."

7. Taha is an epithet of the prophet Muhammad formed from the two Arabic letters Ta and Ha.

8. See Martin 1989.

9. Ibid., 237.

10. Reynolds 1995, chap. 3, "Poets Inside and Outside the Epic."

11. A remarkable incident from southern Egypt in which a poet was not able to disclaim an offending verse is analyzed in detail in Slyomovics 1987.

12. See Burke 1931.

1. Special thanks to Rajmohan Singh Rangad and his family for sharing their home in Gaichwan village with me in 1993-1994. Thanks also to Terry Austrin, Jacob Neusner, Jane Simpson, and Alf Hiltebeitel for their comments on a previous draft. The research upon which this essay is based was made possible by a grant from the University of Canterbury.

2. Chandola, 1977, 18.

3. The critical edition (nineteen volumes) was edited by V. S. Sukthankar (19331966). The only complete English translation is that of K. M. Ganguli, first published by P. C. Roy in 1883-1896 and recently reissued (1981-1982).J. A. B. van Buitenen translated two volumes before his death, and a team of scholars is carrying on with what will certainly become the standard English translation (1973-). The most useful condensed English translation is still that of C. V. Narasimhan (1965).

4. Pollock 1986, 15.

5. Hiltebeitel 1976, 1988.

6. Biardeau 1976, 1978.

7. An invaluable study relating Hindu pilgrimage places to the Mahabharata is Bhardwaj 1973. Other useful overviews of Hindu pilgrimage include Bharati 1970, Eck 1981, and Morinis 1984. Among a plethora of recent studies of Ayodhya and its continuing religio-political importance are Embree 1990 and van der Veer 1987.

8. Galey 1986.

9. For the concept of lila, see Haberman 1988; Hein 1972, 1987; and especially Sax 1994.

10. Egnor 1980; Fuller 1992; Sax 1991b; Waghorne and Cutler 1985.

11. Blackburn and Flueckiger (1989) have also argued that in India, epics are defined in terms of "the extent and intensity of a folklore community's identification with them; they help to shape a community's self-identity" (p. 6); and that "most important, oral epics in India have that special ability to tell a community's own story and thus help create and maintain that community's self-identity" (p. 11). Flueckiger's essay in this volume also shows how a folk epic changes along with the characteristics of its community.

12. The arrowheads are not always used. Since they must receive blood sacrifice, they are not used in villages where such sacrifice is prohibited. In addition, they are not common in western Garhwal.

13. After receiving the bow and arrow from a dancer representing the Pandavas' guru Drona, one holds both of them above one's head, looks up, and spins around while shaking them. Several small bells are attached to the bow, so that the sound of their jingling mixes with the pounding of the drums. There in the firelit square, under the bemused gaze of several hundred villagers, I had the distinct feeling that it was not I who was spinning the weapons around, but they were twirling me.

14. See Sax 1997.

15. Desai 1970.

16. Sukthankar 1933, cii.

17. Nautiyal 1971, 133-134; Sharma 1977, 79; Thukral 1987, 41.

18. This character is not found in the Pune edition of the epic. There is, however, a figure named Vikarna, one of the ninety-nine younger brothers of Duryodhana, who speaks up on Draupadi's behalf during her humiliation in the assembly. Possibly Vikhasan has some relation to him.

19. Polyandry is found in the Jaunsar and Jaunpur regions, a few hours' journey away. See Majumdar 1963.

20. For discussions of bride-price in general, see Tambiah 1973; for bride-price in the Central Himalayas, see Fanger 1987; Sax 1991a.

21. According to one of the god's priests in Jakhol, there are a total of fourteen temples dedicated to the god in the immediate vicinity. I have been unable to confirm reports of a Duhshasana temple.

22. There is a Duryodhana temple in Kerala with a priest who, when possessed by Duryodhana, dances on one leg for several hours (Tarabout 1986, 223, 483; cited in Hiltebeitel 1991, 178). I thank Alf Hiltebeitel for this reference.

23. This "mini-Mahabharata" is the subject of a long folk song that I have transcribed and translated into standard Hindi with the help of Bhuli Das of Dyora village and hope to publish soon.

24. Nautiyal (1971, 133-134) cites the Mahabharata to this effect, but I am unable to verify his claim.

25. Saklani 1987, 44, 174ff.

26. Ibbetson and MacLagan 1919; Raha 1979; Rosser 1955; Sax 199 a; Sutherland 1988.

27. Duryodhana is said to be a partial incarnation of the demon Kali or "discord."

28. jai ho, durijodhan maharaj, dasi durijodhan falane das jakhol, devata ka, somesvar devata ka dasi, jai ho devata somesor, teri maya.

29 In 1991 I had to walk over ten kilometers and cross a major river to reach Jakhol from Sankari. In 1994, I took the twice-daily taxi service.

1. Quotations from Plato are from Edith Hamilton and Huntington Cairns, eds., The Collected Dialogues of Plato (New York: Pantheon, 1961). In this edition, the Ion is translated by Lane Cooper, the Republic by Paul Shorey, and the Theaetetus by F. M. Cornford.

2. Herington (1985) supports this argument by the observation that ''one has only to recite Homer aloud, before a real or imagined throng of people, to reexperience the emotions described by Ion" (p. 13). It is true that we have relatively little information concerning the response of Greek theater audiences to the performances of tragedies, but one early indication is suggestive. Herodotus records an instance in which the tragedy performed produced so many tears that its author was punished: "The Athenians. . . showed themselves beyond measure afflicted at the fall of Miletus, in many ways expressing their sympathy, and especially by their treatment of Phrynichus, for when this poet brought out upon the stage his drama, The Capture of Miletus, the whole theatre burst into tears, and the people sentenced him to pay a fine of a thousand drachmas, for recalling to them their own misfortunes. They likewise made a law, that no one should ever again exhibit that piece" (Herodotus 6.21; trans. G. Rawlinson [New York: Modern Library, 1942]).

3. Odyssey 1.353; trans. R. Fitzgerald (New York: Vintage, 1990).

4. Murnaghan 1987, 180.

5. Trans. R. Lattimore (Chicago: University of Chicago Press, 1961).

6. Nagy 1979, 77-78.

7. My translation.

8. Trans. M. Armour (New York: Dutton, Everyman's Library, 1934).

9. Trans. V. Nabokov (New York: Random House, 1960).

10. See Sháhnáma of Firdausi, pt. 3, secs. 23-25; I refer to the translation of A. G. Warner and E. Warner (London: Kegan Paul, Trench, Tribner and Co., 1906).

11. Plato and Aristotle make analogous remarks elsewhere. Plato states at Theaetetus 152e that all agree on "the supreme masters in each of the two kinds of poetry: Epicharmus in comedy, Homer in tragedy." Aristotle writes in Rhetoric 1403b: "It was long before [the proper method of delivery] found a way into the arts of tragic drama and epic recitations: at first poets acted their tragedies themselves." For these quotations from Plato and Aristotle, I am indebted to Herington 1985, 213-214.

12. Trans. M. G. Kovacs (Stanford: Stanford University Press, 1989), pp. 106-107. For the meaning of the name Gilgamesh, see page xxvii in the introduction to Kovacs's translation.

13. I refer to the translation of F. P. Magoun, Jr. (Cambridge, Mass.: Harvard University Press, 1963), pp. 278-279.

14. It is true that Hesiod placed stress on the power of poetry to bring "forgetfulness of evils and a rest from cares" ( Theogony 55; see also lines 98-103). But this description has to be interpreted in the light of the pain and discord contained in Hesiod's own poetry.

15. Russo and Simon 1968, 492.

16. My attention was first drawn to the suggestive ambiguity of this line by Marion Wells.

17. Trans. C. Singleton (Princeton: Princeton University Press, 1973).

18. Quint 1983, 115.

1. For this approach, see Martin 1989, 44.

2. Bowra 1952, 10.

3. On this legislation, see Alexiou 1974, 14-23; Holst-Warhaft 1992 114-119.

4. On the epitaphios logos as representing a deliberate rejection of the lament, see Loraux 1986, 42-50.

5. Holst-Warhaft (1992, 127-170) interprets tragedy as an appropriation and denigration of women's laments. Foley (1992) paints a more ambiguous picture, arguing that tragedy registers both the danger of lament and the authenticity of the issues it raises.

6. Alexiou 1974, 11-14.

7. On grief ( penthos ) as antithetical to kleos and on the audience's noninvolvement in the story as an essential element in the realization of kleos, see Nagy 1979, 95-100.

8. For this aspect of Homeric poetics, see Ford 1992.

9. On the laments of the Iliad as occasions when women emerge as commentators on the events of the poem, see Easterling 1991.

10. On this speech, see further Murnaghan 1992, 262-263. On the widespread association of birdsong with mourning, see Alexiou 1974, 97; and Feld 1982.

11. For mourning as the occasion of similar competition among women in modern Inner Mani, see Seremetakis 1991, 89-92, and pp. 130-144 for a case in which a woman makes use of lament to disclaim rather than to claim a relationship with a dead man (she has been engaged to him but wants to remain free for another marriage). For non-Greek depictions of women competing as mourners, see Ovid Amores 3.9.53-56; Holst-Warhaft 1992, 7.

12. As Nagy (1979, 111) points out.

13. Martin 1989, 87-88.

14. See Martin 1989, esp. 222-223, for this characterization of Achilles. Martin places Achilles' affinity for lament within the context of his deployment of the larger heroic speech genre of recollection, classifying lament as a version of recollection (pp. 131, 144-145). Although Martin is surely right that lament's connection to future fame also allies it with other commemorative genres (see p. 86), the classification of lament as simply a form of recollection overlooks the range of its themes, which include fantasy and speculation about the future as well as memories of the past, and obscures the degree to which a man who laments is using a mode of speech that is primarily feminine and antiheroic.

15. For a detailed reading of the relationship between Achilles' and Briseis's laments that stresses Achilles' marginalization, see Pucci 1993.

16. Alexiou 1974, 171-175.

17. Loraux 1990, esp. 77-100. On the relationship between that figure and the lamenting mothers of Homeric poetry, such as Hecuba, Thetis, and Penelope, see Murnaghan 1992.

18. For the expression, see Kakridis 1949, 18-27, and, for its application to this episode, pp. 49-53. Kakridis identifies the motif as a link between this episode and the story of Meleager, the narrative in which a wife successfully uses the language of lament to persuade her husband to fight to defend her (although there as an alternative to not fighting at all rather than, as in Hector's case, fighting too aggressively).

19. See also Iliad 24.748, 762.

20. Martin 1989, 136-137.

21. Andromache seems no more concerned with Hector's kleos as a source of reflected glory for herself than as a compensation to him for his loss of life. She herself expresses no interest that would be met by the consideration raised by G. S. Kirk (1990, 222) in his attempt to soften the element of exploitation in Hector's words at 6.460-461: "His reaction to Andromakhe's imagined fate might seem strangely self-centered; that would be typically heroic, but Hektor knows she will be remembered mainly through himself." In general, Homeric women in their laments represent themselves as losing the kind of status conferred by fame, stressing the annihilating displacement suffered by women whose male defenders are gone. Cf. Penelope in the Odyssey, who twice explicitly declares that her kleos has been compromised by Odysseus's absence ( Odyssey 18.251-255, 19.124-128). Cf. also Iliad 22.431-436, discussed below.

22. As Holst-Warhaft (1992, 112-113) points out, none of the three women who lament at Hector's funeral praises him as a hero in battle.

23. On the distinction between kleos as the term for fame that transcends mortality and its near doublets, kudos and euchos, which is equivalent to euchole, see Muellner 1976, 82, Io.

24. Alexiou 1974, 165-171.

25. Sultan (1991) gives a suggestive account of modern Greek Akritic song as a genre in which, similarly, women's interests are opposed to male heroic action and yet women's voices are essential to the process of making heroic glory immortal.

1. Alexiou 1971. See the discussion of the kommos (antiphonal lament) in Fantham 1982, 219-231.

2. Scholars have noted the possible influence of the Greek tragic chorus on the unattributed speeches of communal lament or protest in Roman epic from Ennius to Lucan and Statius.

3. Both terms denote the advocate's final emotional address, but while miseratio properly denotes the appeal to pity for his client, and conquestio the provocation of anger against the adversary, or a third party, these two functions often overlap.

4. Frag. lxi, ed. Skutsch: "O Romule Romule Die/ qualem te patriae custodem di genuerunt!/ 0 pater, O genitor, O sanguen dis oriundum!/ Tu produxisti nos intra luminis oras."

5. Livy 2.7.5 records that the married women mourned Brutus for a whole year because he had been so keen an avenger of woman's violated honor (the rape of Lucretia).

6. Lucan De Bello Civili 7.37-39. The laments are described as unbidden (virtually forbidden) because Caesar will be in absolute power over Rome.

7. Iustitium: Lucan De Bello Civili 2.18. On precedents for this decree, see Fantham 1992, 83.

8. See Lucan De Bello Civili 2.21-28 (comparison), 29-36 (behavior of lamenting women).

9. Livy 22.55.3-8: the women's laments for their war dead were so disturbing that by public decree they were confined to their homes and forbidden to make further public outcry.

10. "The vaulted walls echo with the wail and woe of women: the matrons wander, clutching at the doors/ embracing them, imprinting kisses" ("penitusque cavae plangoribus aedes / femineis ululant; ferit aurea sidera clamor. / tum pavidae tectis matres ingentibus errant/ amplexaeque tenent postis atque oscula figunt," Aeneid 2.487f., trans. Mandelbaum).

11. "The lamentations, keening, shrieks of women/ sound through the house, heavens echo mighty wailings" ("lamentis gemituque et femineo ululatu/ tecta fremunt, resonat magnis plangoribus aether," Aeneid 4.666-667, trans. Mandelbaum).

12. Wiltshire 1989, esp. 52-53 (in chap. 2, "Grieving Mothers and the Cost of Attachment").

13. "Wretched she runs out, and with a woman's wailing, tearing her hair,. . .  she fills heaven with her cries" ("femineo ululatu/ scissa comam . . . caelum dehinc questibus implet," Aeneid 9.477-480, trans. Mandelbaum). This is followed by fifteen lines of her lament in direct speech.

14. Hardie (1993, 49) comments: "Euryalus' mother is in fact entirely selfish in her desire for death"; but he acknowledges the influence of both Euryalus's own demand to die instead of his friend and his mother's demand for death on the Virgin's Lament in Renaissance Christian epics.

15. On Juturna's lament, see Barchiesi 1978.

16. "Quis deus, O Musae, paribus tot funera verbis/ evolvat? tantisque umbris in carmine digna/ quis lamenta ferat?" Punica 5.420-422.

17. Until recently there was little secondary literature on the Thebaid in English (or indeed other languages). Readers should consult Vessey 1973, which has a full bibliography of previous work; see also Ahl et al. 1986 and most recently Hardie 1993.

18. Besides the consolationes for the bereavements of Statius's friends, Silvae 2.6 and 3.3, see Statius's Epikedeia (Laments) for the wife of Abascantus, for his own father, and for his adopted child, Silvae 5. , 3, and 5.

19. I quote from Henderson 1991, his first and more complex paper.

20. Omitted in his English diagram: see Henderson 1991, 31. The Latin equivalence of lament is represented by eight forms: ei mihi, heu, gem-, plang-, quer-, lament-, fle-, dol-/ O. Three ( ei mihi, heu, 0! ) are cries of grief; the others are roots of descriptive nouns and verbs.

21. Henderson 1991, 78 n. 191.

22. Ibid. 60.

23. Thebaid 3.98-99, 102-103, trans. Melville. A. D. Melville's translation of the Thebaid, the first since J. H. Mozley's rather inaccurate Loeb version, may strike the American reader as too archaic, and so convey a more artificial impression of Statius's diction than his admittedly heightened epic language.

24. Makarismos, the blessing invoked on the honored dead (such as Virgil's blessing on Nisus and Euryalus in Aeneid 9 ), is only a specialized form of the Greek collective epitaphios logos, for those who died in battle or the Roman laudatio, enumerating a man's virtues and achievements at his funeral.

25. Virgil is more sparing of personification but includes Grief with Avenging Cares ( Luctus et Ultrices . . . Curae ) among the phantasms that crowd the entrance to the underworld in Aeneid 6.274. As we shall see below, grief and vengeance are clearly and repeatedly associated in Statius.

26. Like the heroine of Sergei Eisenstein's Alexander Nevsky, but in that optimistic epic the heroine finds both her brother and fiancé alive.

27. The progressive articulation of response to sorrow /death between women's mourning (for the past) and men's call to revenge (for the future) echoes the contrast noted above between Euryalus's mother and Evander, or even between Lucan's matrons and warriors in book 2. On the Maeon episode and the role of Aletes, see also Ahl et al. 1986, 2830-2831 n. 19.

28. On the provocation of resentment against the Athenian generals, see Xenophon Hellenica 1.7.8; against the defeated Servilius Caepio, see Cicero De Oratore 2.201.

29. See further note 31 below.

30. Menoeceus is in fact singled out for this sacrificial death because he is the youngest descendant of the serpent-born, earth-begotten warriors who peopled Cadmus's city of Thebes. See further Vessey 1971.

31. Grief and weeping are stressed here too; cf. Thebaid 12.26, 32, and the miserabile certamen of 33-34, disputing claims over the dead.

32. Contrast Hardie 1993, 48: "The epic ends not with triumph but with lament, or rather with the praeteritio of lament, a programme to inspire another epic (808 novus furor ) to balance the programme at 1.33-45 for the epic we have already read." It will be clear that I see lament itself as a triumph-of a new kind.

33. Just before Parthenopaeus's death the goddess Diana, unable to prevent it, had addressed the boy: "You delight in the battles so bewailed" ("ululataque proelia gaudes, " Thebaid 9.724). Statius's application of ululare, the vocabulary of lament, to battle is new and almost programmatic. Diana's next words ("happy, and dying only for your mother," "felix et miserae tantum moriture parenti," Thebaid 9.725) thematically polarize male delight in battle and parental lament. But at the end that "only'' is corrected by Statius's own celebration of the mother's grief for "her son,/ Her son who kept his grace though blood was gone,/ Her son for whom two armies grieved as one"( Thebaid 12.806-807, trans. Melville).

34. See Ahl 1986, 2905.

35. I would like to express my warm thanks to Randall Ganiban for his valuable insights and suggestions, which have enriched and improved the written revision of this essay.

1. Swann 1994 provides an important recent translation of Native American literature; see also Elder and Wong 1994. For more information on the storytelling traditions of the Labrador Indians, see Millman 1992, 1990, and 1987.

2. Throughout this essay, I will be referring to R. D. Williams's edition of Virgil's Aeneid and to Robert Fitzgerald's translation (New York: Random House, Vintage Books, 1983). This essay benefited from a careful reading by Denis Feeney, who is of course not to be blamed for its limitations.

3. This and future references to the Odyssey in English will be to Robert Fitzgerald's translation (Garden City, N.Y: Doubleday, Anchor Books, 1961).

4. See Wofford 1992a, 136-146.

5. On the regress of narratives characteristic of efforts to define origin, see Guillory 1993, esp. 23-45.

6. See Quint 1993, 9 and passim, including especially pp. 248-267 (in chap. 6, "Tasso, Milton and the Boat of Romance"); and the opening chapter of Bellamy 1992.

7. See Hardie 1993, 1-18, on the way that epics strive for completion and totality but are never successful in reaching either.

8. See Williams 1972, 1: 395, 433.

9. I have relied on A. D. Melville's translation of Ovid's Metamorphoses (Oxford: Oxford University Press, 1986). For the Latin, I have used the Loeb edition, except for passages from books 6-10, for which I have used Anderson's edition and commentary (1972).

10. See Forbes Irving 1990 for a typology of early Greek myths that tell of metamorphoses. This account divides the transformation stories into those about (a) mammals; (b) birds; (c) plants; (d) stones; (e) springs and rivers; (f) islands; and (g) insects, reptiles and sea-creatures. Forbes Irving finds distinct story patterns to be typical of each kind of metamorphosis. Most important for my purposes here, Forbes Irving distinguishes between those metamorphoses that serve as aetiologies for a species and those that create one unique, often monstrous representative.

11. See Stinton 1976, 60-89, an account of this problem and the use of similar phrases ("they say") in classical literature. On the question of this kind of fictive truth claims in the ancient world, see Feeney 1993, 230-244.

12. See Forbes Irving 1990 for a catalogue of bird transformations involving various kinds of pollution, almost all sexual in some way or another. Forbes Irving sees bird transformation stories in Greek myth as being principally structured around an opposition between the house and the wilds, and the stories of pollution tend, therefore, to be stories that begin with a "family crime" (p. 107)-incest of various sorts, with related family murder; a few examples of sex with animals. "But these polluting crimes can generally be seen as outbreaks of wildness," Forbes Irving writes, "and it is the more general opposition of the wilds and human order that is the basic structure of these stories" (p. 109). See Forbes Irving, pp. 107-109.

13. The question of the extent to which metamorphosis tends to produce an unchanging nature-the species existing in the world as we know it, no longer subject to a metamorphic principle-is discussed with complexity in Barkan 1986. Barkan argues that Ovid's metamorphoses, at least, function to initiate the reader into an understanding of "the metaphoric flow among separate categories of existence" (p. 31), and yet he also notes that some metamorphoses are "changes to end all change," and comments that "the natural world is one endpoint of transformation" (p. 79). In treating the changes that end all change, though, Barkan focuses mostly on the Roman aetiologies in Ovid's poem, showing how the poem tends to undermine covertly its proclaimed story that transformations all lead to the creation of Rome and Roman institutions. Throughout his influential account of Ovid, Barkan is interested in the tension between the metamorphic principle—which can at times wreak havoc in a human context, as when categories that define social order dissolve and the principle of fixity, by which the individual is fixed in an essentialized and often "savagely reduced'' version of his or her identity (p. 66). Barkan's account of the basic structure of an Ovidian metamorphosis is as follows:

[The story] takes place in a context where individuals are assigned clear roles, so clear that they may be oppressive. The central figure in the story rebels, specifically attacking the clarity and discreteness of the surrounding categories. The essential metamorphosis comes as a direct result of this rebellion: it is not the hero's or heroine's change of physical shape . . . but rather the discovery that what seemed like rigid categories of family and society can dissolve, just as physical categories dissolve in metamorphosis. Once the categories are attacked, similar things are diversified into opposites and opposites are made identical. The central figure reaches a condition that transcends and contradicts all these categories. From that point it is a short step to literal metamorphosis, a condition that merely serves as a final punctuation mark for a narrative experience whose crucial metamorphosis has amounted to the dissolution of assumptions we live by (p. 59).

Barkan's emphasis on the dissolution of ordinary assumptions about differences between, say, the human and the wild, or between the sexes, is salutary, and an essential feature of stories of origin is precisely that the narrative events that precede the creation of the world as we know it tend to involve transgressions of that world. It is still useful to note, though, how that "final punctuation mark" serves even in Barkan's metamorphic theory as a stabilizing and fixing closure to the story.

14. See Myers 1994 for a discussion of Ovid's poetry in relation to the genre of the aetion, and especially Myers's useful account of the way the first and last of Ovid's amatory tales serve as a framing device for the narrative.

15. On the frequency with which trees are used as comparisons for human growth and beauty in the metamorphoses associated with Greek myth, see Forbes Irving 1990, 134.

16. My understanding of the account of Daphne and Apollo has been immeasurably increased by having studied the story with Denis Feeney. See Feeney 1991, 205-224; see also p. 216 on the associations of Apollo and Augustus, and p. 219: "The emphasis on the constitutional facts of Roman cult has its corollary in the politically coloured anthropomorphism which infuses the epic descriptions of deities in action. . . . The preoccupation with such systems of analogy is something we have been familiar with since the actions centering on the poem's first simile, in which Jupiter was compared to Augustus." Feeney is discussing the more extended analogy in the poem between Augustus and Jupiter, but part of his point is that the absolute, irrational, and tyrannical behavior of the gods may often comment on what imperial power felt like to those on whom it was exercised.

17. See Barkan 1986, 66, for an account of how a metamorphosis is "an escape from entrapment into a higher condition where the blurred categories are no longer meaningful."

18. For the tension between elegy and epic in Ovid's two accounts of the rape of Persephone, where the origins of the world as we know it are located specifically in rape, see Hinds 1987, 115-134. On the frequency of rape in accounts of origins, see Wofford 1992b and Jed 1989.

19. From the note in the Melville translation, p. 383.

20. On the Greek sources of the Philomela story, see Forbes Irving 1990, 248-249. Forbes Irving comments on the extent to which in many of these accounts the transformation tends to be a form of escape, even when the existing species is marked or signed with the events of the narrative: "Since they are describing every day facts of nature and since the present birds are no longer really the original transformed people these birds will tend to be less alarming than the unique freaks created by transformation into an animal; also the positions of these transformations at the end of their story, and the fact that they will last forever, removes any sense of urgency about the heroes' behaviour or state of mind as birds, and produces a perhaps artificial sense of serenity about their new state. They are now free from any further human tragedy, in comparison with which neither the suffering of birds nor any further evil that they do among each other needs to be taken too seriously" (p. 112). See also Barkan 1986, 65, on the way in which for criminals against the natural order (such as Tereus) metamorphosis is "not a punishment but rather a definition of the extreme state into which they have brought themselves and a relief from the agony of those extremes."

21. Anderson 1972, 236.

22. For all references to "The First Loon," see the appendix, where the story is reproduced in full, and see Millman 1992.

23. Many indigenous origin tales do include precisely the kinds of detailed partto-part analogy that impress readers of Ovid (her arms became branches, etc.). See, for one example that involves extensive analogy, the Tahitian tale "Tangaroa, Maker of All Things":

"But his anger was not finished, and so he took his backbone for a mountain range and his ribs for the ridges that ascend. . . . He took his fingernails and toenails for the scales and shells of fishes in the sea. Of his feathers he made trees and shrubs to clothe the land." (Elder and Wong 1994, 40)

24. As Millman (1992) reminds us in his note to the story, Michikimau, the setting of the ending of the story, was a lake sacred to the Innu, which vanished in 1963 when the Churchill Falls Hydro Project was constructed. The tale is a tale of loss in more than one sense.

25. For all references to "The Origin of Robins," see the appendix, where the story is reproduced in full, and see Milliman 1992. This tale was told to Millman by Thomas Pastitshi, Utshimassits.

26. See Puttenham 1970, bk. 3, chap. 19, pp. 236-238.

27. See Murnaghan 1987 for a discussion of how the romance motifs such as this rejuvenation reverse epic temporality.

28. See Wofford 1992a, 112-113, for a more extended discussion of exactly how much Aeneas sees here and what he knows.

29. See Feeney 1991, 48-49, on the anthropomorphism of the gods in Homeric commentaries, and pp. 134-137 on the way that this anthropomorphism in the Aeneid signals that the gods are fictive. Feeney observes that "the gods of the Aeneid may be 'figures', 'tropes, 'symbols', but only as Aeneas is a figure or a trope. Neptune and Juno need the same fictive energy to be creators and channellers of meaning as does Aeneas. If we accept that the gods are figural, symbolic, in that they refer to areas of meaning beyond themselves, we need not then feel baffled at their 'unnecessary' colour and force" (pp. 136-137). In a private communciation, Feeney also noted that "the anthropomorphic representation of angry deities had been criticized and scrutinized for centuries—it is not an ultimate explanation or cause but a metaphor always."

30. On the centrality of the repetition compulsion in Virgil and Renaissance epic, see also Bellamy 1992, a psychoanalytic study of the problem of origins in the epic tradition.

31. All quotations from Spenser's Faerie Queene are from A. C. Hamilton's edition (London and New York: Longman, 1977). See Guillory 1983 for Spenser's use of fountains as markers of an origin that has always been literary; and Quint 1983 for the topos of the fountain or spring as image for literary source. On the connections of Spenser's fountain to Ovid's aetiologies in the myth of Hermaphroditus and elsewhere in the Metamorphoses, see also Silberman 1987.

32. Contrast the opening of book 4 of Lucretius's De Rerum Natura: "A pathless country of the Pierides I traverse, where no other foot has ever trod. I love to approach virgin springs [intregros . . . fontis], and there to drink." Thanks to Denis Feeney for this reference.

33. For an account of this problem see Wofford 1992a, 262-281 (the section entitled "The Cause Was This").

34. I necessarily (because of limitations of scope in one essay) skip over here the Christian tradition of allegorizing metamorphosis, and most notably Dante, but we might briefly note that the metamorphoses in the Inferno (for instance, in the Wood of the Suicide) are also irreversible, a critical problem in the poem (where the fact that the damned cannot be saved is understood as a way of explaining how the sinners in the Inferno repeatedly and eternally make again the sinful choice that caused them to lose human form in the first place).

1. For a brief bibliography and survey of the critical tradition, see Marowski and Matuz 1987, 414-423.

2. In this paper I cite from Walcott 1986 ( Collected Poems, 1948-1984 ) unless otherwise noted. The passage in question may be found on p. 15. The poem originally appeared in 1956, according to Irma E. Goldstraw's indispensible bibliography (1984, 5), and was subsequently included in Walcott 1962, 18.

3. Walcott 1986, 443. The poem originally appeared in The New Yorker in 1981 (Goldstraw 1984, 39) and, later that year, in the collection The Fortunate Traveller ( New York, 1981), 63-70.

4. Lefkowitz 1090, 1,34-35; Knox 1991,3-4; Taplin 1991,213-226; Steiner 1993, 13-16.

5. Burris 1991, 559 (my emphasis).

6. Hamner 1993b, 19; cf. the introductory remarks in Hamner 1993a, 10-12.

7. Figueroa 1991, 211.

8. Ismond 1991, 10-11.

9. Quoted by Bruckner 1990, reprinted. in Hamner 1993a, 396-399. Walcott's remarks stress the importance to him of the novelists Rudyard Kipling, Joseph Conrad, and Ernest Hemingway as models for Omeros. See also Brown and Johnson 1990, 209-233; White 1990, 14-37.

10. "Omeros is profoundly Homeric and undoubtedly epic"; see Taplin 1991, 213-226.

11. Burris 1991, 560.

12. On dramatic elements in Omeros, see Burris 1991, 561-564. Burris calls Ulysses "the work that will in all likelihood emerge as the most generous sponsor of Omeros " (p. 561).

13. Figueroa 1991, 203-205. On p. 205 he observes that St. Lucy, the patron saint of the island, was herself a blind seer. Blindness and compensatory insight is a recurring theme in Walcott's work, one with special relevance to the figure of the poet. In "Cul de Sac Valley" the poet images himself as an Oedipus questioned by a row of Sphinxes (Walcott 1986, 13).

In this essay I will use the capitalized form "Narrator" to indicate the character in Omeros who narrates the poem and represents the figure of the poet himself; the lowercase form indicates the implied singer of whatever poem happens to be under discussion.

14. Figueroa 1991, 206.

15. Leithauser 1991 Cf. Figueroa 199  197: "The poem [is] much more a novel than an epic, while never losing its lyrical fire."

16. The argument is spelled out most clearly in Bakhtin's essay "Epic and Novel" (Bakhtin 1981, 3-40).

17. The question of Walcott's influences, which has been prominent in criticism of his work since the beginning, came to be viewed in terms of cultural allegiance as Walcott's European influences were found by some less relevant to the Africanist West Indian political consciousness of the 1960s and 1970s than the work of other writers, particularly Edward Brathwaite. The literature comparing the two writers is quite large: representative works include Lucie-Smith 1968; Drayton 1970; James 1970; [Anonymous] "How Far are Derek Walcott and Edward Brathwaite Similar?" 1974; Collier 1979; King 1980. As Walcott's interest in African themes, particularly in plays such as Dream on Monkey Mountain and O Babylon!, came to be appreciated, the question of his cultural allegiances became less urgent. Further, with Walcott's rise to international stature he has come to be compared with poets such as Joseph Brodsky and Seamus Heaney, and one result of appearing in such company before an international audience is that his Caribbean identity seems hardly in doubt. Significantly, the West Indian writer with whom he is most often contrasted nowadays is not Brathwaite, but V. S. Naipaul, with whose dismal judgment upon postcolonial culture, particularly in the West Indies, Walcott (1974) took exception.

18. Bowra 1952, 1-11.

19. Ibid.

20. Finnegan 1970, 108ff. Finnegan, however, is not concerned, as Bowra is, with the capacity of Africans to produce heroic literature so much as with the technical question of whether their heroic literature is in verse.

21. Of crucial importance was the publication of the Sundiata epic (Niane 1960; Pickett 1965 (English translation). On the poem, see Miller 1990, 87-101. Other important scholarly investigations of African epic include Okpewho 1979; Knappert 1983; William Johnson 1980.

22. On the work of Parry and Lord on this tradition, see Lord 1960.

23. Kambili, vol. 1 of The Songs of Seydou Camara, trans. Charles S. Bird, Mamadou Koita, and Bourama Soumaoro (lines 505-507); cited by Okpewho 1979, 205-206.

24. A more apt comparison might have been between other instances of interaction between poet and audience in contemporary performative epic and passages in our Iliad and Odyssey that are best explained as "local variants," that is, as versions of the story suited to performance in some specific setting that somehow found their way into what eventually became the "canonical" text. Such an explanation has been advanced in the case of the episode involving Aeneas in Iliad 20, which may ultimately owe its existence to a ruling dynasty that claimed descent from the hero: see Kirk 1991, 298-301, with further references. An even stronger case can be made for the prominence of the Athenian contingent in the Catalogue of Ships in that Athens was not a great power either at the time when the events of the Iliad putatively occurred or at the time when a recognizable version of the poem was first coming into existence; yet the Athenian tyrant Peisistratus played some role, one that may have been both extensive and decisive, in the canonization of the text of Homer that has come down to us. On this particular problem see Kirk 1985, 178-180, with further references. On the phenomenon in general, see Svenbro 1976, 5-73.

25. This passage has a long history of interpretation, much of which finds the humorous element misplaced. See Kirk 1990, 190-191, with further references.

26. With this motif we may compare contemporary performances of North African epic: see the Dwight Reynolds's essay in this volume: "Problematic Performances: Overlapping Genres and Levels of Participation in Arabic Oral Epic-Singing."

27. Walcott 1987, 10. 

28. Walcott 1978, 155-156. This passage is quoted at greater length and discussed in Terada 1992, 93-94.

29. Seven Seas performs, for instance, at a party held at the café in honor of a political candidate (2.20.1). Ma Kilman's eventual role as Philoctete's healer underlines the assonance between her name and that of Machaon, surgeon to the Greek forces in the Iliad, as Burris points out (1991, 561), citing the equivalence as an example of Walcott's "slapstick disregard" for his Homeric parallels. Burris's rather facile reaction ignores the fact that the character of Ma Kilman, a "gardeuse, sybil, obeah-woman" ( 1.10.2), antedates Omeros and indeed is first presented not as Walcott's creation, but as "found object" of St. Lucian folk culture, appearing first in a Creole song included and translated in "Sainte Lucie," secs. 4-5 (Walcott 1986, 314-319, first published in the collection Sea Grapes as long ago as 1976). The connection with Machaon would appear to have been forged or ''discovered" some time after the poet's initial acquaintance with the figure. I would add that the hand of the poet is more clearly visible in the character's connection with the No Pain Cafe, which takes its name from that of, "[allowing] no pain," a drug administered by Helen to her husband, Menelaus, and to their guests, Telemachus and Peisistratus, so that they might discuss the war at Troy and the difficult homecomings of the Greeks who fought there without succumbing to grief. Thus Ma Kilman herself is a type of Helen in her odyssean, as opposed to her iliadic, manifestation.

30. Omeros 7.56.3.

31. The theme of alleged gaps in the author's reading recurs, again with respect to the sources of Omeros, but this time involving the Aeneid as well as the Odyssey, in White 1990, 16-35. The problem is addressed with great insight by Fuller (1992 517-538). One thinks of Yeats's striking way of naming the inspiration of his life's work: "the half-read wisdom of daemonic images" ("Meditations in Time of Civil War" 7.40, in Yeats 1983, 206).

32. Omeros 7.56.3.

33. The theme of a natural language heard or even read in landscape is prominent throughout Walcott's work. See for instance the excellent observations of Terada 1992, 152, 164-165, 167, 171-174.

34. The locus classicus for this line of discourse is Friedrich Schiller's essay Über naive und sentimentalische Dichtung ( 1795-1796 ). It continues in G. W. F. Hegel's Asthetik of 1835 (on which see Bowie 1990, 140-142), Georg Lukács's Theorie des Romans (1920), Erich Auerbach's Mimesis ( 1953 ) and Bakhtin's "Epic and Novel" (1941; not widely known in this country before the Emerson and Holquist translation of 1981). It is fair to say that the influence of these thinkers on the study of the novel and its relationship not only to epic but to premodern literature in general has been decisive, but in many ways far from constructive.

35. Figueroa 1991, 21 1; cf. Omeros 7.56.3: [Narrator] "The gods and the demigods aren't much use to us."/"Forget the gods," Omeros growled, "and read the rest."

36. On ancient allegoresis of Homer, see Lamberton 1986, with further references. On Paradise Lost, see Newlyn 1993. For a convenient survey of twentieth-century trends in Virgilian criticism, see Harrison 1990, 1-20.

37. This particular type of intertextuality goes by the convenient name oppositio in imitando. There is a considerable literature on this phenomenon, most of it known, unfortunately, only to specialists. For a brief survey with references, see the introduction to Farrell 1991, 3-25. As a convenient illustration of the effect produced by this type of writing, consider the Narrator's observation that Achille's "end, when it comes, will be a death by water/(which is not for this book)" ( Omeros 7.64.1). The point being imitated is Tiresias's prophecy in the Homeric Odyssey that the hero's death will occur far from the sea ( Odyssey 11.134-136). The imitation e contrario not only redefines the meaning of death at sea according to the values of a new poetic universe but actively enlists the contribution of a whole range of previous independent imitators of the Odyssey, from Dante, whose Ulisse does in fact contradict Homer by dying a watery death ( Inferno 26.85-142), to Kazantzakis, whose importance to Walcott as a mediator of Homeric and meta-Homeric traditions awaits further exploration, and Eliot, particularly of course in The Waste Land, to mention only these.

The phenomenon of oppositio in imitando parallels what Harold Bloom has famously figured as the belated poet's struggle for originality in the face of an oppressive weight of tradition in The Anxiety of Influence (1973) and subsequent studies, but differs by focusing largely on the impersonal forces of generic development rather than on the psychological trope of the Oedipus complex. A further parallel may be found in the work of those scholars who have attempted to define the role of the individual poet-singer working within a tradition of oral composition and performance: e.g., Nagler 1974; Austin 1975.

38. On this aspect of the Aeneid, see Putnam 1988, 151-201. On Milton's Christian revision of pagan heroism, see Fish 1967.

39. See Quint 1993; Wofford 1992. See the review of both Quint and Wofford in Farrell 1993.

40. The poem thus privileges orality over literacy: the modern spelling is identical to the ancient, but the rough breathing mark is vestigial since the initial Hsound has disappeared (Walcott uses the H only at 3.30.2). Thus Walcott's transliteration of Homer's Greek name into Roman characters as "Omeros" ironically represents more accurately than standard modern Greek orthography not only the absence of the H sound, but also the fact that the first and second 0 sounds (represented in Greek by omicron and omega, respectively) no longer differ in quantity, as they did in the ancient language, but actually sound identical. In fact, to say even this is too simple in view of the multiplicity of ancient conventions of spelling and pronunciation and the modern distinction between Katharevousa and Demotike. But my main point is, I think, clear.

41. Note that it is clearly an inhabited or personified environment: a conch shell sounds only when blown like a horn; leaves may crunch under human footsteps or from other causes; and the mouth of the cave quickly becomes the Narrator's mouth.

42. Terada 1992, 60. "Ruins of a Great House" originally appeared in 1956, then in Walcott 1962 (the collection In a Green Night: Poems, 1945-1960 ) and most recently in Walcott 1986 ( Collected Poems ) 19-21.

43. Omeros 5.41.1.

44. Omeros 5.41.2-3.

45. Omeros 3.27.1.

46. "Buffalo Soldiers" was the name given by the Southwest and Plains Indians to the troops who served between 1866 and 1891 as the Ninth and Tenth Regiments of the United States Cavalry, all of them African-Americans. The troops evidently accepted the name as a badge of honor, and the Tenth incorporated a bison into its regimental emblem (Leckie 1967, 25-26). The Ninth's involvement in the U. S. government's response to the Ghost Dance movement among the Sioux in 1890-1891 was the last significant campaign of the Buffalo Soldiers (Leckie, pp. 25-26). The narrative of this episode in Omeros occurs in what may be the most elliptical part of the poem. It begins when Achille, fresh from his hallucinatory voyage to Africa, remembers hearing the Bob Marley song "Buffalo Soldier" at a party the previous night and imagines himself a member of that troop ( Omeros 3.31.1). The tale is related sporadically in the thirteen chapters that flow through the Narrator's experiences living in Boston and, especially, traveling to the Great Plains (a trip explicitly likened to Achille's dream of Africa at 4.34.2) and in passages related from the perspective of Catherine Weldon, a Boston woman who lived with Sitting Bull at the time of the Ghost Dance. This thread of narrative ends with book 5. Achille himself, in the reverie induced in him by Marley's music, is imagistically associated with the destruction of the Sioux nation and of the Aruacs (3.31.1).

In a similar way, Achille's ancestor, the Afolabe who first acquired the name Achilles from Admiral Rodney himself, helped the British forces position a cannon for the defense of St. Lucia against a French assault (2.14.3). By this act this Achilles unwittingly takes the part of the British Empire, which would ultimately gain political control over the island, against the nation that would leave so great a stamp on the island's culture, particularly its language and religion, in the time of his descendants.

It is possible, although Omeros does not do so, to document the converse phenomenon, the complicity of American Indians in the enslavement of blacks: see Abel 1992. And, to complete this brief typological survey of racial oppression, see Koger 1985.

47. Omeros 5.41.2.

48. "Now that you are twice my age, which is the boy's/which the father's?" "Sir,"—I swallowed—''they are one voice" (1.12.1).

49. Bloom 1973.