¹ Lord, Albert B., The Singer of Tales (Cambridge, Mass., 1960Google Scholar; repr. New York, 1968, etc.); and Magoun, F. P., ‘Oral-Formulaic Character of Anglo-Saxon Narrative Poetry’, Speculum 28 (1953), 446–67CrossRefGoogle Scholar, repr., e.g., The Beowulf Poet, ed. Fry, Donald K. (Englewood Cliffs, N.J., 1968), pp. 83–113Google Scholar, whose pagination is cited herein. For a complete history of formulaic theory in Old English, see my ‘Introduction: the Oral Theory in Context’, Oral Traditional Literature: a Festschrift for Albert Bates Lord, ed. Foley, John Miles (Columbus, Ohio, 1981; repr. 1983), pp. 51–122Google Scholar. For more general accounts, see Haymes, Edward R., A Bibliography of Studies Relating to Parry's and Lord's Oral Theory (Cambridge, Mass., 1973)Google Scholar; and Foley, , ‘Oral Literature: Premises and Problems’, Choice 18 (1980), 487–96Google Scholar; and idem, Oral-Formulaic Theory: an Introduction and Annotated Bibliography (New York, 1984).Google Scholar

² For nineteenth-century interest in patterned diction, see ‘The Oral Theory in Context’, pp. 51–9.

³ This is Parry's original definition, given in ‘Studies in the Epic Technique of Oral Verse-Making. 1. Homer and Homeric Style’, repr. The Making of Homeric Verse: the Collected Papers of Milman Parry, ed. Parry, Adam (Oxford, 1971), p. 272Google Scholar; I cite this volume hereafter as MHV.

⁴ As an example of the synchronic approach see Nagler, Michael N., Spontaneity and Tradition: a Study in the Oral Art of Homer (Berkeley, California, 1974)Google Scholar; as an example of diachronic analysis see Nagy, Gregory, Comparative Studies in Greek and Indic Meter (Cambridge, Mass., 1974)CrossRefGoogle Scholar; and idem, The Best of the Achaeans: Concepts of the Hero in Archaic Greek Poetry (Baltimore, Maryland, 1979).Google Scholar

⁵ See, e.g., Arant, Patricia, ‘Formulaic Style and the Russian Bylina’, Indiana Slavic Stud. 4 (1967), 7–51Google Scholar; and Foley, , ‘Tradition-Dependent and -Independent Features in Oral Literature: a Comparative View of the Formula’, Oral Traditional Literature, ed. Foley, pp. 262–81.Google Scholar

⁶ Parry, Milman, L'Epithète traditionnelle dans Homère: Essai sur un problème de style homérique (Paris, 1928)Google Scholar, trans. Parry, Adam as ‘The Traditional Epithet in Homer’, MHV, pp. 1–190Google Scholar; and idem, Les Formules et la métrique d'Homère (Paris, 1928)Google Scholar, trans. Parry, Adam as ‘Homeric Formulae and Homeric Metre’, MHV, pp. 191–239.Google Scholar

⁷ Two early investigations of formulaic structure and aesthetics were Calhoun's, George M. ‘Homeric Repetitions’, Univ. of California Publ. in Classical Philol. 12 (1933), 1–25Google Scholar; and idem, ‘The Art of Formula in Homer – eпea пtepoenta’, Classical Philol. 30 (1935), 215–27, to, 215–27Google Scholar, to which latter Parry replied in ‘About Winged Words’, Classical Philol. 32 (1937), 59–63Google Scholar, repr. MHV, pp. 414–18.

⁸ Ed. Fenik, Bernard C. (Leiden, 1978).Google Scholar

⁹ See, e.g., Kirk, G. S., The Songs of Homer (Cambridge, 1962)Google Scholar; and Homer and Oral Tradition, ed. Kirk, (Cambridge, 1976)Google Scholar; Parry, Adam, ‘Have We Homer's Iliad?’, Yale Classical Stud. 20 (1966), 177–216Google Scholar; Parry, Anne, Blameless Aegisthus: a Study of amymΩn and other Homeric Epithets (Leiden, 1973)Google Scholar; and Russo, J. A., ‘Homer against his Tradition’, Arion 7 (1968), 275–95Google Scholar. As Fenik puts the quandary (‘Stylization and Variety: Four Monologues in the Iliad’, Homer: Tradition and Invention, ed. Fenik, p. 90)Google Scholar: ‘How shall we account for the not infrequent cases of carelessness and imprecision in Homer, set as they are against others where he demonstrates mastery of detail, conceptual penetration and architectural design? The ruling theory of the day (oral poetry) explains only half. The sheer quality of Homer remains an unicum in the body of oral poetry known to us.’

¹⁰ With evidence of both oral traditional structure and literary craftsmanship, many a critic would agree with Fenik (‘Homer and Writing’, Würzburger Jahrbücher für die Altertumwissenschaft n.s. 2 (1976), 47)Google Scholar that ‘the dividing line between Mündlichkeit and Schriftlichkeit remains one that we still cannot accurately draw’ [italics deleted].

¹¹ ‘Oral–Formulaic Character’, p. 84.

¹² Cf. Creed, Robert P., ‘The Making of an Anglo-Saxon Poem’, Eng. Lit. Hist. 26 (1959), 445–54Google Scholar, repr., with ‘Additional Remarks’, The Beowulf Poet, ed. Fry, pp. 141–55Google Scholar; and Greenfield, Stanley B., ‘Grendel's Approach to Heorot: Syntax and Poetry’, Old English Poetry: Fifteen Essays, ed. Creed, Robert P. (Providence, Rhode Island, 1967), pp. 275–84.Google Scholar

¹³ ‘The Canons of Old English Criticism’, Eng. Lit. Hist. 34 (1967), 141–55.Google Scholar

¹⁴ PMLA 81 (1966), 334–41.Google Scholar

¹⁵ Apart from Parry's brief mention of traditional narrative elements (‘On Typical Scenes in Homer’, Classical Philol. 31 (1936), 357–60Google Scholar, repr. MHV, pp. 404–7, itself a review of Walter Arend's Die typischen Scenen bei Homer), the first description of typical scenes as oral traditional elements was Lord's ‘Composition by Theme in Homer and South Slavic Epos’, Trans. of the Amer. Philol. Assoc. 82 (1951), 71–80Google Scholar; for earlier articles which touch on related aspects of oral composition, see his ‘Homer and Huso’ series, Trans. of the Amer. Philol. Assoc. 67 (1936), 106–13Google Scholar, 69 (1938), 439–45 and 79 (1948), 113–24. For fuller expositions, see his The Singer of Tales, where he defines themes as ‘groups of ideas regularly used in telling a tale in the formulaic style of traditional song’ (p. 68); and idem, ‘Perspectives on Recent Work on Oral Literature’, Forum for Mod. Lang. Stud. 10 (1974), 187–210Google Scholar, repr. Oral Literature: Seven Essays, ed. Duggan, Joseph J. (Edinburgh and London, 1975), pp. 1–24.Google Scholar See, further, my ‘Introduction’, pp. 79–122.

¹⁶ ‘Literary Character’, p. 337.

¹⁷ As Fry puts it (‘Cædmon as a Formulaic Poet’, Forum for Mod. Lang. Stud. 10 (1974), 228–48Google Scholar, repr. Oral Literature: Seven Essays, ed. Duggan, pp. 41–61, at 41)Google Scholar: ‘a consensus seems to be emerging that written Old English poetry used oral forms, but no reliable test can differentiate written from oral poems’.

¹⁸ ‘Literary Character’, p. 340.

¹⁹ See, esp., Fry, ‘Cædmon’, and Lord, ‘Perspectives’. Payne, Richard C. (‘Formulaic Poetry in Old English and its Backgrounds’, Stud. in Med. Culture 11 (1977), 49)Google Scholar argues that ‘given the evidence of deliberate composition and balanced literary design in Old English poems that many investigators of the problem have produced, it seems likely that most poems were produced by authors with pen in hand, though frequent communal reading of such works must be assumed to maintain the vitality of the formulaic tradition’.

²⁰ See my ‘Formula and Theme in Old English Poetry’, Oral Literature and the Formula, ed. Stolz, Benjamin A. and Shannon, Richard S. (Ann Arbor, Michigan, 1976), pp. 207–32Google Scholar. Of related interest are Niles, John D., ‘Formula and Formulaic System in Beowulf’, Oral Traditional Literature, ed. Foley, pp. 394–415Google Scholar; and Foley, , ‘Beowulf and Traditional Narrative Song: the Potential and Limits of Comparison’, Old English Literature in Context: Ten Essays, ed. by Niles, John D. (London and Totowa, N.J., 1980), pp. 117–36 and 173–8.Google Scholar

²¹ A bibliography of writings on oral-formulaic theory published by Garland Press (see above, n. 1) contains more than 1800 items in dozens of literatures, most of which appeared after the publication of The Singer of Tales in 1960.

²² Much new work is appearing each year from traditions all over the world: a telling example is Görög's, Veronika enormous Littérature orale d' Afrique Noire: Bibliographie analytique (Paris, 1981).Google Scholar

²³ Hrvatske narodne pjesme, ed. Marjanović, Luka, iii–iv (Zagreb, 1898)Google Scholar; Narodne pjesme Muslimana u Bosni i Hercegovini, ed. Hörmann, Kosta, 2 vols. (Zagreb, 1888–1889Google Scholar; repr. Sarajevo, 1976); and Serbo-Croatian Heroic Songs (Srpskohrvatske junačke pjesme), collected, ed. and trans. Parry, Milman, Lord, Albert and Bynum, David E., i–iv, vi and xiv (Cambridge, Mass., and Belgrade, 1953–1980).Google Scholar

²⁴ ibid. iii–iv.

²⁵ See Lord, , ‘The Theme of the Withdrawn Hero in Serbo-Croatian Oral Epic’, Prilozi za knjiźevnost, jezik, istoriju i folklor 35 (1969), 18–30Google Scholar; and Foley, , ‘The Traditional Structure of Ibro Bašić's “Alagić Alija and Velagić Selim”’, Slavic and East European Jnl 22 (1978), 1–14CrossRefGoogle Scholar. For a bibliography of writings on the Return Song see Foley, ‘Beowulf and Traditional Narrative Song’, Old English Literature in Context, ed. Niles, p. 173Google Scholar, n. 6, to which should be added Coote, Mary P., ‘Lying in Passages’, Canadian–Amer. Slavic Stud. 15 (1981), 5–23CrossRefGoogle Scholar, and W. W. Parks, ‘Generic Identity and the Guest–Host Exchange: a Study of Return Songs in the Homeric and Serbo-Croatian Traditions’, ibid. pp. 24–41.

²⁶ On some alternative quantitative methods for approaching these problems, see Miletich, John S., ‘Oral–Traditional Style and Learned Literature: a New Perspective’, Poetics and the Theory of Lit. 3 (1978), 345–56.Google Scholar

²⁷ This flexibility is one aim of much of Jeff Opland's scholarship; see, esp., ‘Imbongi Nezibongo: the Xhosa Tribal Poet and the Contemporary Poetic Tradition’, PMLA 90 (1975), 185–208Google Scholar, and Anglo-Saxon Oral Poetry: a Study of the Traditions (New Haven, Conn., 1980).Google Scholar

²⁸ Generic taxonomy is ever a matter of dispute in dealing with Serbo-Croatian oral poetry, and much of the Vuk material may just as convincingly be labelled ‘lyric’ on the basis of its brevity. I choose, however, to maintain the label of ‘epic’, the term favoured by many native investigators in spite of the songs’ length, because of their epic subjects and phraseology.

²⁹ On the manuscripts, see Mladenović, Živomir, ‘Rukopisi narodnih pesama Vukove zbirke i njihovo izdavanje’, Srpske narodne pjesme iz neobjavljenih rukopisa Vuka Stef. Karadžića 1, ed. Mladenović, Živomir and Vladan, Nedić (Belgrade, 1973), i–cclxxix.Google Scholar

³⁰ ‘Rukopisi narodnih pesama’, p. iv (translation mine). See, further, Mladenović, ‘Vuk kao redaktor narodnih pesama’, Kovčežić 1 (1958), 64–93.Google Scholar

³¹ Vuk's editing amounted to one substantive change (a hemistich substitution or an added line) approximately every seventy-five lines. Most changes involved lesser matters, such as spelling, dialect forms and morphology, as explained below.

³² Quotations from Vuk, with the exception of examples cited directly from the manuscripts, are taken from the edition of Vladan Nedić (Belgrade, 1969). ‘Djakon Stefan i dva andjela’ is preserved in part (lines 1–19) as br. 8552/257, xxii, 5 of the Archive of the Srpska Akademija Nauka i Umetnosti in Belgrade.

³³ ‘Ljuba Jakšića Šćepana’, archive br. 8552/257, vii, 57.

³⁴ In the standard edition of Srpske narodne pjesme, ed. Aleksić, P. et al. ii (Belgrade, 1958)Google Scholar, in the ‘Beleške i objašnjena’, p. 734, we find the following remark: ‘This song is not truly heroic, but more “midway between heroic and lyric [women's] songs” (according to Vuk's distinction), like those in volume one. Vuk included it here because it tells of Kraljević Marko’ (translation mine).

³⁵ See, in vol. ii of Vuk's collection, ‘Marko Kraljević i Mina od Kostura’ 193–4, where the drinking of wine acts as a restorative. On the context provided to an oral poem by its tradition, see Caraveli, Anna, ‘The Song Beyond the Song: Aesthetics and Social Interaction in Greek Folksong’, Jnl of Amer. Folklore 95 (1982), 129–58.CrossRefGoogle Scholar

³⁶ See, e.g., Whitman, Cedric H., Homer and the Heroic Tradition (Cambridge, Mass., 1958CrossRefGoogle Scholar; repr. New York, 1965), esp. pp. 249–84; John Niles, D., ‘Ring Composition and the Structure of Beowulf’, PMLA 94 (1979), 924–35.CrossRefGoogle Scholar

³⁷ Čelo gori (‘the forehead burns’) is a figurative way of indicating either shame, and therefore respect, before one's superior or impatience or possibly even anger. Both readings are supported by the passage and together present the dilemma of divided allegiance which Marko must continually confront as a Christian hero and yet a Turkish vassal and mercenary.

³⁸ In Homeric Formulae and Homeric Metre, Parry explains various metrical flaws in the hexameter as the juxtaposition of formulas which, although each is metrically admissible by itself, create a flaw at their point of juncture. An example is the instance of short-vowel hiatus in ἔνθa κaθέζεт' ἔπειта Ὀδυσσ⋯οѕ фίλоѕ υἱόѕ (Odyssey xvi.48), which he derives as a variation on the same model which produced ἔνθa κaθέζεт' ἔπειта περίфρων Πηνελόπειа(Odyssey xix. 59), among others (p. 203). Inconsistencies at the level of the scene can result from an analogous juxtaposition of themes which by themselves are internally consistent, but which when placed alongside one another do not agree in certain details; see, further, Lord, , ‘Homer and Huso 11: Narrative Inconsistencies in Homer and Oral Poetry,’ Trans. of the Amer. Philol. Assoc. 69 (1938), 439–45Google Scholar; idem, The Singer of Tales, pp. 68–98.

³⁹ See, further, Foley, , ‘Narrativity in Beowulf, the Odyssey, and the Serbo-Croatian Return Song’, Proceedings of the IXth Congress of the International Comparative Literature Association, Innsbruck 1979. 1: Classical Models in Literature, ed. Konstantinović, Zoran, Anderson, Warren and Dietze, Walter, Innsbrucker Beiträge zur Kulturwissenschaft, Sonderheft 49 (Innsbruck, 1981), 295–301.Google Scholar

⁴⁰ See, further, Foley, , ‘Epic and Charm in Old English and Serbo-Croatian Oral Tradition’, Comparative Criticism: Yearbook of the Brit. Comparative Lit. Assoc. 2 (1980), 71–92, esp. 76–7.Google Scholar

⁴¹ For a deftly drawn portrait of Podrugović see Koljević, Svetozar, The Epic in the Making (Oxford, 1980), pp. 311–14Google Scholar (this volume is a revision and expansion of his Naš junčki ep (Belgrade, 1974))Google Scholar; also relevant are pp. 117–23. On p. 311 Koljević, quoting periodically from Vuk, remarks: ‘“He was clever and, for an outlaw, an honest man”; and it was at this time, when Podrugović was about forty, that Karadžić recorded twenty-two poems from him. “I have never found anyone who knew the poems as well as he did. Each of his poems was a good one, because he – particularly as he did not sing but spoke his poems – understood and felt them, and he thought about what he said.” He had a large repertoire and knew, in Karadžić's opinion, at least another hundred poems apart from the recorded ones; moreover, Karadžić claims that if Podrugović were “to hear the worst poem, after a few days he would speak it beautifully and in proper order which was characteristic of his other songs, or he would not remember it at all, and he would say that it was silly, not worth remembering or telling”.’

⁴² ‘The Effect of the Turkish Conquest on Balkan Oral Tradition’, Aspects of the Balkans: Continuity and Change, ed. Birnbaum, Henrik and Vryonis, Speros Jr (The Hague and Paris, 1972), esp. pp. 311–18Google Scholar. See also above, n. 25.

⁴³ In narrative terms, this is the result of the collision between the Return Song pattern, which must always end in a wedding or a reunion of husband and wife (a remarriage, from the folkloric viewpoint), and the particular case of Marko, for whom marriage never seems possible; see, further, in vol. ii of the Vuk collection, “Ženidba Kraljevića Marka” (‘The Wedding of Kraljevic Marko’).

⁴⁴ Cf., in vol. ii, the champion Miloš in ‘Ženidba Dušanova’ (‘The Wedding of Dušan’).

⁴⁵ See above, n. 15.

⁴⁶ Mejdandžije, translated here as ‘combatants’, literally means ‘those who fight a duel’, usually but not always as necessary replacements for their sovereigns, much as Gawain takes up the gauntlet for King Arthur in Sir Gawain and the Green Knight. The mejdan (‘duel’) itself is a common motif in many songs and customarily determines the disposition of disputed property or the betrothal of a maiden.

⁴⁷ Koljević describes Višnjić's achievement as ‘the richest evidence of the interplay of the formulaic, the formulopoeic and the unique’ (The Epic in the Making, p. 340), indicating that this singer was able to combine stock epic phraseology and narrative design both with unique and with derivative language and scenes. See, further, his discussion of this blend of tradition and innovation in ‘Technique and Achievement’, ibid. pp. 322–43, the final section of which applies directly to Višnjić's songs.

⁴⁸ Compare the obvious parallel at Beowulf 677–87, where the hero prepares for the battle against Grendel by disarming himself, a variation made more powerful by the traditional arming sequence which it reverses.

⁴⁹ It is important to recall at this point that the Muslim songs should not be criticized for their apparent lack of concern with such aesthetic issues as those discussed here, any more than the briefer poems should be taken to task for lack of development. The longer epic is simply a poetic form sui generis and operates on its own principles, as the writings of Lord and others have so well illustrated. We are dealing with a matter of genre, that is, with a taxonomic distinction: what is possible in a shorter and perhaps memorizable form is not possible in the much longer epic composed under the pressure of oral performance. The distinction is not, to put it another way, one of quality; it is one of genre.

⁵⁰ Speculum 35 (1960), 87–96Google Scholar. Campbell allows minor morphological variants and substitution of function words within his conception of the formula.

⁵¹ ‘Another Look at Oral Poetry in The Seafarer’, Speculum 35 (1960), 596–600.Google Scholar

⁵² ‘Oral Poetry in The Seafarer’, p. 91, n. 10.

⁵³ It may be worth noting that a poet who can consciously manipulate traditional style to aesthetic advantage has presumably freed himself from outright dependence on formulaic structure and so has disqualified himself from eligibility for the formulaic test. Cf. Russom, Geoffrey R., ‘Artful Avoidance of the Useful Phrase in Beowulf, The Battle of Maldon, and Fates of the Apostles’, SP 75 (1978), 371–90Google Scholar; Haymes, Edward R., ‘Formulaic Density and Bishop Njegoš’, Comparative Lit. 32 (1980), 390–401.CrossRefGoogle Scholar

⁵⁴ All quotations of Old English poetry are from the collective edition of George P. Krapp and E. V. K. Dobbie, The Anglo-Saxon Poetic Records, 6 vols. (New York, 1931–53). Supporting evidence is drawn from A Concordance to the Anglo-Saxon Poetic Records, ed. Bessinger, Jess B., Jr, prog. Philip H. Smith (Ithaca, N.Y., and London, 1978)Google Scholar, with abbreviations as used therein: soðgied wrecað (Vgl 15b); gebiden hæfdon (Exo 238b), gebiden hæbbe (DrR 50b), gebiden hæbbe (DrR 79b) and gebiden hæbbe (Bwf 1928b); atol yða gewealc (Exo 456a); and of nacan stefne (And 291b).

⁵⁵ On this aspect of phraseology, see Niles, John D., ‘Compound Diction and the Style of Beowulf’, ES 62 (1981), 489–503.Google Scholar

⁵⁶ In ‘Old English Formulas and Systems’, ES 48 (1967), 193–204Google Scholar, Fry discusses the formulaic system, which he defines as ‘a group of half-lines, usually loosely related metrically and semantically, which are related in form by the identical relative placement of two elements, one a variable word or element of a compound usually supplying the alliteration, and the other a constant word or element of a compound, with approximately the same distribution of non-stressed elements’ (203; italics deleted), and the formula itself, which he then defines as ‘a group of words, one half-line in length, which shows evidence of being the direct product of a formulaic system’ (204). See, further, above, n. 17.

⁵⁷ It is important to emphasize that this particular type of multiformity is a tradition-dependent property of Old English verse. As I have shown elsewhere (see above, n. 5), the Homeric Greek and Serbo-Croatian prosodies also prescribe idiosyncratic features in the phraseologies they support.

⁵⁸ Parry first described the operation of sound-patterning in oral traditional composition, calling it a form of ‘analogy’ (e.g., ‘The Traditional Epithet in Homer’, MHV, pp. 68–75)Google Scholar. See, further, Lord, , ‘The Role of Sound-Patterns in Serbo-Croatian Epic’, For Roman Jakobson, ed. Morris, Halle et al. (The Hague, 1956), pp. 301–5Google Scholar; idem, The Singer of Tales, pp. 54–8; Stanford, W. B., ‘Euphonic Reasons for the Choice of Homeric Formulae?’, Hermathena 108 (1969), 14–17Google Scholar; Creed, , ‘The Beowulf-Poet: Master of Sound-Patterning’, Oral Traditional Literature, ed. Foley, pp. 194–216.Google Scholar

⁵⁹ This is no less true of Serbo-Croatian and ancient Greek epic, in which traditional composition depends to varying degrees on exactly repeated phrases, flexible systems and phonetic patterns, all of which take shape under the aegis of prosody.

⁶⁰ See, further, Greenfield, , ‘The Formulaic Expression of the Theme of “Exile” in Anglo-Saxon Poetry’, Speculum 30 (1955), 200–6.CrossRefGoogle Scholar

⁶¹ On verbal echo see the early work of Beaty, John O., ‘The Echo-Word in Beowulf with a Note on the Finnsburg Fragment’, PMLA 49 (1934), 365–73CrossRefGoogle Scholar; and later articles by Kintgen, Eugene R.: e.g., ‘Wordplay in The Wanderer’, Neophilologus 59 (1975), 119–27.CrossRefGoogle Scholar

⁶² For The Wife's Lament 1a, cf. Beowulf 2446b: þonne he gyd wrece.

⁶³ ‘The Old English Elegies’, Continuations and Beginnings: Studies in Old English Literature, ed. Stanley, Eric G. (London and Edinburgh, 1966), p. 147Google Scholar. The other words involved in this design are dream (65a, 80a and 86b), blæd (79b and 88b), duguð (80a and 86a) and lof(73a and 78a).

⁶⁴ Line 65a; dryhtnes dreamas (Glc 123a); line 121b: þa for lufan dryhtnes (Ele 491b), þæt hie lufan dryhtnes (Ele 1205 b), ond mid lufan dryhtnes (Glc 652b), þæt hi lufan dryhtnes (Jul 501b) and ond a lufan dryhtnes (Jgl 49b); and line 124a: ece drihten (Gen 112b) and eighty-seven other occurrences in the poetic corpus. Of the three remaining lines, all are verse-types beginning with unstressed elements, that is, those types most susceptible to changes through substitution of function words; in addition, lines 41a and 106a are hypermetric, the former also being part of a rhetorical series. Although it is beyond the scope of this article to discuss the phenomenon at any length, parallels to various kinds of rhetorical series in The Seafarer can be located in the Vuk songs, so that we need not suppose such stylized figures the exclusive creation of the post-traditional poet.

⁶⁵ As an example of the controversy over formulaic structure and aesthetics, see Parry, Anne Amory, ‘The Gates of Horn and Ivory’, Yale Classical Stud. 20 (1966), 3–57Google Scholar, versus Lord, , ‘Homer as Oral Poet’, Harvard Stud. in Classical Philol. 72 (1968), 1–46.CrossRefGoogle Scholar

⁶⁶ ‘The Formulaic Expression of the Theme of “Exile”’, p. 201.

⁶⁷ Ibid., esp. pp. 201–5.

⁶⁸ Most investigators have sought to define Old English themes in terms of a recurring narrative pattern rather than actual verbal correspondence among instances. See, e.g., Crowne, David (‘The Hero on the Beach: an Example of Composition by Theme in Anglo-Saxon Poetry’, NM 61 (1960), 362–72)Google Scholar, who defines this typical scene as ‘a stereotyped way of describing (1) a hero on the beach (2) with his retainers (3) in the presence of a flashing light (4) as a journey is completed (or begun)’ (368) and notes that ‘the regular content of this theme, then, consists not of a number of specific metrical formulas, but of a concatenation of four imagistic details’ (371). See, further, my review of scholarship on the theme: Foley, ‘Introduction’, pp. 79–122.

⁶⁹ ‘Beowulf on the Poet’, MS 38 (1976), 445–53.Google Scholar

⁷⁰ On the roots of words as verbal correspondence among instances of a theme see Foley, ‘Formula and Theme in Old English Poetry’, pp. 220–32; idem, ‘Beowulf and Traditional Narrative Song’, pp. 122–34; Lord, ‘Perspectives’, pp. 20–4; Peabody, Berkley, The Winged Word: a Study in the Technique of Ancient Greek Oral Composition as Seen Principally through Hesiod's ‘Works and Days’ (Albany, N.Y., 1975)Google Scholar, on ‘The Responsions of Thought’, pp. 195–215. It is worth noting in this connection that not all themes in Serbo-Croatian answer precisely the same definition in terms of formulaic content and particular kinds of structure (see, e.g., Coote, Mary P., ‘The Singer's Themes in Serbocroatian Heroic Song’, California Slavic Stud. 11 (1980), 201–35)Google Scholar, and so perhaps we should not expect all Old English themes to exhibit exactly the same features.

⁷¹ Compare, for example, the panegyric material described by Opland (see above, n. 27) or the many shorter genres noted by Ruth Finnegan in her Oral Literature in Africa (Oxford, 1970)Google Scholar. On the importance of matching genre in comparing oral poetries see Foley, , ‘Oral Texts, Traditional Texts: Poetics and Critical Methods’, Canadian–Amer. Slavic Stud. 15 (1981), 122–45.CrossRefGoogle Scholar

⁷² See, esp., Lord, (‘Memory, Fixity, and Genre in Oral Traditional Poetries’, Oral Traditional Literature, ed. Foley, pp. 459–61)Google Scholar, who notes that ‘since memory may play a somewhat different role in different genres, it seems necessary to take genre into account in discussing fixity’ (p. 459). Cf. Holoka, James P. (‘The Oral Formula and Anglo-Saxon Elegy: some Misgivings’, Neophilologus 60 (1976), 570–6)CrossRefGoogle Scholar, who argues that ‘in a short work … the factor of improvisation is much less operative. Certainly a skilled singer, one who could run on for thousands of verses without (we must suppose) serious breakdown, could easily enough hold in his mind, in toto, a set piece of some one hundred verses; he could review, polish, revise, rework until finally his method closely approximated that of his more educated counterpart. Thus, short, elegiac poems could conceivably attain a fixity indistinguishable from that of a written text’ (572).

⁷³ It is a pleasure to express my appreciation to the institutions that supported the research for this essay and the individuals who contributed to its development. The International Research and Exchanges Board, affiliated with the American Council of Learned Societies, together with the Yugoslav Federal Administration for International Scientific, Educational, Cultural and Technical Cooperation, provided a stipend for five months of research in Belgrade. Both the Odsek za jugoslovenske književnosti of the Filološki fakultet at Belgrade University and the staff of the Archive of the Srpska Akademija Nauka i Umetnosti were very gracious and helpful, the latter in allowing me direct access to the original manuscripts of Vuk Karadžić's collection. The first version of the essay was written during a 1980–1 leave sponsored by the John Simon Guggenheim Memorial Foundation, and sections were later presented to sessions on Old English poetry at the 1981 Medieval Institute and Modern Language Association. I am especially grateful to Professor Nada Milošević-Djordjević for her extensive discussions of the Vuk material.