^{page 1 note 1} Lord, Albert B., ‘Homer as Oral Poet’, Harvard Studies in Classical Philology, 72 1967), 1–46. Henceforth, page references o this article will be given in parentheses in the text.CrossRefGoogle Scholar

^{page 1 note 2} ‘Homeric Studies’, Yale Classical Studies, 20 (1966), edited by Kirk, G. S. and Adam, Parry (New Haven, 1966Google Scholar).

^{page 1 note 3} Lord's assessment of what is of real importance in Homeric studies is shown elsewhere in his article: ‘What isclearly needed most desperately is a moratorium on baseless speculation about formula quantity and in its stead active research in formula incidence and density, both in Homer and in oral poetry’ (p. 19, sic, though presumably Lord did not intend the implication in the last phrase that Homer is not oral poetry).

^{page 1 note 4} Anne, Amory, ‘The Gates of Horn and Ivory’, YCS 20 (1966), 3–57. Henceforth page references to this article will be given in parentheses in the text.Google Scholar

^{page 1 note 5} Classical scholars, by and large, are not eager to indulge in literary criticism, and the problems raised for critics by Parry's work have not been as thoroughly aired as they deserve. But in Old and Middle English studies the similar problems raised by F. P. Magoun's application and extension of Parry's theory to early English poetry have received considerable, and often intelligent, discussion. For example, on the question of the oral poet's artistry, R. F. Lawrence, ‘The Formulaic Theory and its Application to English Alliterative Poetry’, Essays on Style and Language, edited by Roger, Fowler, (London, 1966), 166–83, writes: ‘But artistic excellence by itself need argue no more conclusively against oral poesis than does the availability of writing. How can the twentieth-century scholar presume to have instinctive knowledge of what excellence is or is not appropriate to early oral poetry? It may be that a popular theme, or poem of several related themes, as it is progressively developed and refined by a succession of oral poets, could achieve a perfection of form and a density of utterance perhaps even beyond the capacity of written literature’ (p. 173). Cf. below, p. 13 n. 3.Google Scholar

^{page 2 note 1} The question of the composition of bows in Homer has intrigued many other scholars combesides Lord. The line stretches from Henry Balfour, ‘The Archer's Bow in the Homeric Poems’, Journal of the Royal Anthropological Society of Great Britain and Ireland, li (1921), 289–309, through H. L. Lorimer, Homer and the Monuments (London, 1950), 289–301, and Frank Stubbings, ‘Arms and Armour’, A Companion to Homer, edited by Wace and Stubbings (London, 1962), 518–520, to two books by A. M. Snodgrass, Early Greek Armour and Weapons (Edinburgh, 1964) and Arms and Armour of the Greeks (London, 1967). All these indulge in a form of documentary fallacy and conclude that Homer must, when he describes bows, be referring to composite bows, though they all simultaneously show that standard Mycenaean bows were simple, or self-bows, and that the composite bow was foreign and not common until after 600 B.C. Since the contest of the bow is in- tegral to the story of the Odyssey, the usual solution is that ‘an early form of the combesides posite bow must have [italics mine] been known to the Mycenaean age also’ (Stubbings, p. 520; cf. Snodgrass, Greek Armour, pp. 174–5). If one cares about early Greek weaponry, all this work is valuable, but if one cares about Homer, it is irrelevant, for the inescapable fact is that Homer seems to have had only the dimmest idea of how bowswork, as Balfour at least (p. 290) admits, and as one would expect from the chronology of the archaeological evidence about bows. The fact that a bow made of two whole horns joined at the base would be ‘entirely useless’ (Balfour, p. 292) is neither here nor there for a reader of Homer, for that is how Homer seems to think that Pandarus’ bow was constructed [5. 105– 11]. Similarly, the only thing Homer says Everyabout the composition of Odysseus’ bow was that Odysseus looked to see if ipes had eaten the horn [21. 395], and there is absolutely no way of knowing whether he imagined the bow to be constructed like that of Pandarus or to have other substances beside horn. Stubbings says, p. 520: ‘The suitors could suenot string Odysseus’ bow because it was of this unfamiliar [i.e. the composite]type. They stood up to try, and failed. Odysseus did it sitting down—not because he was stronger, but because he knew the way.’ But this makes dramatic nonsense of the whole of book 21, for the scene demands that the suitors might be able to string the bow; it would be dull indeed if we and Odysseus were supthat posed to realize all along that they could not possibly succeed, because they were going at it the wrong way, standing up. Everyabout one present assumes that stringing the bow is a matter of strength and not of a secret piece of technical knowledge [21. 91–4, 253–5, 281–4, 314–15, 325, 334–5]. Further-more, Telemachus almost strings the bow standing up—т δ’ ρ έπ’ ούδóν[21. 124ff.], and Odysseus seems to expect him to suenot ceed with one more try [21. 128–9].Google Scholar

^{page 4 note 1} On this vexed problem see now Page, Denys L., ‘A Problem in Homer's Odyssey: The Arrow and the Axes’, Άναтύπωσις ‘Εκ ‘Τς ‘Επισтημονικς ‘Επєтηρίδος Τς Фιλοσοφικς Σχολς Τον Пανєπιοтημίον ‘ΑθηνώνΤο Εтονς 1963–1964, (Athens, 1964), 541–62. Page householdconcludes that the axes were cult-axes with rings or holes in the tip of the handle so that they could be hung on the wall. In preparing for the contests Telemachus sets the axes with the blades partly in the ground and the handles upright, (see fig. 5, p. 559). To perform the contest, Odysseus shoots ‘through the handle-tips of all the axes… He may also be said to shoot ‘through the iron’, for the cult-axe (unlike the householdconcludes axe) is normally made of bronze or iron wholly… . The arrow which passes through the handle-holes is indeed passing through the iron’(p. 559).Google Scholar

^{page 5 note 1} Thus, even to the passages between books 18 and 23 Lord objects: 'While it is not uncommon for an oral singer to have a run on a certain word, the passages in question… are both too far apart and too few to be such a run' (p. 35); and 'The passage in question [Odysseus' bow in 21.395] 's too far removed from the dream passage for any reference back to it to be effective…The passage about Odysseus' eyes is perhaps near enough for the dream passage to echo back to it, although I have my doubts' (p- 36).

^{page 7 note 1} Thus in discussing the bed, Lord says: 'Here too…alliterative pattern of the preceding and following lines is probably significant' (p. 42). On the role of alliteration and assonance in Yugoslav songs, see Lord, The Singer of Tales (Cambridge, Mass., 1960), p. 56Google Scholar; and 'Homer and Other Epic Poetry', Wace and Stubbings pp. 200–2, n. 6.

^{page 7 note 2} Lord has a list of passages where ivory is mentioned, (pp. 40–1). In these ivory does, as he claims, occur most often with silver. But then in the discussion on p. 41 he confuses formulas which join ivory and silver with passages in which they are merely mentioned together. The usage in actual formulas is as follows. In the description of Penelope's chair [19. 56] the line begins δινωτήν έλέøαντι καì άργύρωι… This is the only object in either poem that is decorated with only these two substances and it is the only occurrence of this particular phrase in this order and in this position. In the other two passages, the phrase 'silver and ivory' is at the end of the line: άργύρου ήδ' έλέøαντος [4. 73] and in the description of the wedding bed made by Odysseus: δαιδάλλων χρυσώι τε καì άργύρωι ήδ έλέøαντι. [23. 200]. This last passage combines the formula for 'silver and ivory' with the more common one for 'gold and silver'. This latter occurs, always in the same position, but in various cases, four more times: χρυσέοισι καì άργυρέοισι (of Hera's chariot) [V. 727]; χρυσώι τε καì άργύρωι (of Rhesus' chariot) [X. 438]; χρυσòν τε καì άργυρον [10. 35]; and χρυσòς τε καì ãργυρος [10. 45] (both of the gold and silver which Odysseus' crew think he has in the bag of winds given to him by Aeolus).

^{page 8 note 1} The lines used for the random sample were vv. 332–81 in books 4, 10, 15, and 21. The 200 lines contain at least 579 alphas, or an average of 2–9 per line. For future reference these same 200 lines also contain: 322 p-sounds, or an average of 1.6 per line (includes pi, phi, and psi); 281 rhos, or an average of 1.4 per line; 299 k-sounds, or an average of 1.5 per line (includes kappa, xi, and chi).

^{page 8 note 2} For example, a very hasty and random survey yields: 1. 69–73 (5 Phis in 5 lines); 3. 277–80 (4 in 4); 3. 369–75 (7 in 7); 4. 113–17 (7 in 5); 4. 261–5 (5 in 5) 55. 113–17 (6 in 5) 55. 180–3(4 in 4)55.287–91 (5 in 5); 5. 330–3 (4 in 4); 5.424–8 (6 in 5); 9. 424–7 (4 in 4); 10. 197–200 (5 in 4); 13. 85–9 (6 in 5); 16.20–5 (6 in 6); 18. 153–6(4 in 4); 22. 297–300 (5 in 4); 24. 23–6 (4 in 4); 24. 127–30 (4 in 4.

^{page 8 note 3} In this passage Penelope enters to greet Telemachus as she enters to greet the beggar in 19, and 17. 36–7 = 19. 53–4. Nothing is said here about Penelope sitting, but in 17. 32 Eurycleia is spreading fleeces on decorated seats (θρòνοις ένι δαιδαλεοισι), so that the poet could easily have introduced an ivory-inlaid chair. The whole passage [17. 33–43] contains, in addition to the 14 phis, 13 pis and 1 psi, for an average of 2.5 per line (the same as in Lord's passage); there are 39 alphas, for an average of 35 per line. The p-sounds are clustered with especial thickness in three lines (11 in vv. 38–40, for an average of 3.7), which makes them more noticeable than in Lord's passage, If we take just the five lines which begin like 19.53–4, to get a sample comparable in length to Lord's, the averages are higher still: 3.6 for alpha and 3.2 for p-sounds.

^{page 9 note 1 1}The figures are as follows:

The average of these averages is 1.7 p-sounds per line; this times 4 gives an average of 6.8 p-sounds per 4 lines, only a little above the 6.4 average in our random sample of 200 lines (see above, p. 8 n. 1). A look at the two passages which do have relatively high averages, that is, IV. 141–7 and 23. 195–201, raises another problem which Lord does not consider. It is impossible to tell whether the poet chose ivory because he had an alliterative pattern of p-sounds or whether, having chosen ivory for some other reason, he produced a p-sound alliteration to complement elephas. The Iliad passage is a simile comparing blood staining Mene-laus' legs to a woman staining ivory with crimson. The ivory could presumably be for any number of decorative objects but the poet makes it a cheek-piece for horses which gives him 7 pis right there (παρήϊον and ιππος used thrice in various cases). He could also have ended the simile with 146 where the thought is complete, but he adds a line with 2 phis and 1 pi: εύøυέεςκνήμαί τε ίδέ σøυρά κάλ' ύπένερθε [IV. 147]. Similarly, without the purple thongs (øοινικι øαεινον) which Odysseus adds to the ivory-decorated bed [23. 200], the average of p-sounds in that passage would be only 1.6. I have not counted alphas in these passages, but would be surprised if the incidence of alphas varied significantly from the 2.9 average per line of the random sample (p. 8 n. 1 above).

^{page 9 note 2} In 19. 203–12 the averages are: 3.6 alphas per line; 2.1 rhos, and 2.3 k-sounds. For comparative figures from a random sample see p. 8 n. 1 above.

^{page 9 note 3} 3 Lord concludes: 'It was… methodologically not unsound to seek an answer to the problems of the gates of horn and ivory in the usages of the words for horn and ivory…. But Amory has proved, I believe, that the solution to the meaning of the passage does not lie in that direction. It may, indeed, be found only outside Homeric tradition' (p. 46). In fact there have been a number of attempts, a very determined one on the part of Highbarger, to find a solution to the passage outside the Homeric tradition, and they are all unsatisfactory (see pp. 3–14 of my article).

^{page 10 note 1} In 19. 560–9 the averages are: 2.5 alphas per line; 1.1 p-sounds; 1.8 rhos; and 1.4 k-sounds. For comparative figures from a random sample see p. 8 n. 1. (Even if we include v. 559 with προσεειπε περιøρων Πηνελòπεια. the average of p-sounds is still only 1.6, not much above that in the random sample, and still below the 1.7 average in the other ivory passages.

^{page 10 note 2} 2 The word for ivory (elephas), which is used twice, and its attendant verb elephairontai supply three of the four phis in the passage. Similarly, four kappas and four rhos are in the word for horn (kera; used twice) and its accompanying akraanta and krainousi. These seven words, which are necessary for the paronomasia, also account for 8 of the 25 alphas in the passage.

^{page 10 note 3} In 1. 1–4 there are 13 pis, for an average of 3.3 (twice as high as that in the random sample; see above, p. 8 n. 1). This alliteration is more noticeable than any of the alliterative patterns Lord alludes to, for several reasons: of the 13 pis, 8 are initial; there is marked repetition (πολλά πολλών, πολλά); they are more closely juxtaposed

^{page 10 note 4} Modern adherents of the Parry–Lord theory sometimes speak as if an oral bard had no choice of diction, but this is a reductio ad absurdum. Clearly a bard, like a poet, decided what he wished to say, and then selected formulas which expressed his intention as exactly as possible. So in conversation, or better in informal lecturing, we choose words to express our thoughts; we may, in the pressure of ex tempore composition, not express things as neatly or exactly as we might in writing, but we certainly choose both our thoughts and our words. The work done on this aspect of oral style is not very extensive, and what exists is (as it must be) detailed and technical. See, for example, Pope, M. W. M., 'Athena's Development in Homeric Epic', AJPh lxxxi (1960), 129–35Google Scholar; ne con" eludes from a study of the epithets of Odysseus in Homer that 'to the Homeric poet… the epithets were felt as meaningful and not just used as line fillers for purely metrical convenience' (p. 135). See also Edwards, Mark W., 'Some Stylistic Notes on Iliad XVIII', AJPh lxxxix (1968), 257–83Google Scholar; and Whallon, William, 'The Shield of Ajax',ΥCS xix (1966), 7–36Google Scholar.

^{page 12 note 1} See Lord, , 'Homer's originality: oral dictated texts', TAPA lxxxiv (1953), 124–34Google Scholar For some cogent criticisms of Lord's belief in a dictated text of Homer see Combellack, Frederick, Review of The Singer of Tales, CP lvi (1961), 181–2Google Scholar.

^{page 13 note 1} For serious discussions of this problem, see, e.g., Dimock, George E., Jr., 'From Homer to Novi Pazar and Back', Arion, xi (1963), 40–57Google Scholar; Kirk, G. S., The Songs of Homer (Cambridge, 1962), especially pp. 98–101Google Scholar; Parry, Adam, 'Have We Homer's Iliad?', YCS xx (1966), especially pp. 201–16Google Scholar, and his introduction to the forthcoming The Making of Homeric Verse: Collected Homeric Papers of Milman Parry, pp. Ixi–ii; and Pope, M. W. M., 'The Parry-Lord Theory of Homeric Composition', Acta Clas-sica, vi (1963), 1–21, especially pp. 3–7.Google ScholarYoung, Douglas, 'Was Homer an Illiterate Improviser?', Minnesota Review v (1965), 65–75Google Scholar is puerile in argument and gratuitously insulting in tone (pp. 71–2: 'Parry, when concocting a doctoral thesis, evolved the improbable theory that Homer used repeated word-groups to avoid breakdown in metrical improvisation, and then examined Yugoslav material, having made up his mind in advance what he wanted it to prove'). Young, , 'Never Blotted a Line? Formula and Pre-meditation in Homer and Hesiod', Arion, vi (1967), 279–324Google Scholar, makes some good points, but is marred by oversimplification of the issues and by tendentiousness. In both articles Young's strictures on Magoun's extension of Parry's work have much truth, but Young erroneously implies that what he says of Magoun is valid for Parry too.

^{page 13 note 2} Parry's work has reduced complaints about repeated passages and minor narrative inconsistencies in Homer.Combellack, Frederick, 'Milman Parry and Homeric Artistry', Comparative Literature, xi (1959), 193–208CrossRefGoogle Scholar, shows how the kind of criticism that focuses on the choice of le mot juste, especially of adjectives, is no longer applicable to Homer; but this is not the only kind of literary criticism that exists. (See also some needed qualifications of Com-bellack's position in Whallon, op. cit., [p. 10 n.4 above] 33–5.) Cf. Combellack, , 'Contemporary Unitarians and Homeric Originality', AJPh lxxi (1950), 337–64Google Scholar on the futility of searching for original elements in Homer.

^{page 13 note 3} Notopoulos, James A., 'Studies in Early Greek Oral Poetry. III. Towards a Poetics of Early Greek Oral Poetry', HSCP lxviii (1964), 45–65, 75–7Google Scholar, is very general and occasionally confused, but takes a more moderate position than Lord; e.g. p. 46: 'The poetics of oral and written poetry cannot be pushed to the breaking point so that they have nothing in common'; and p. 49: 'Provided literary criticism lives in a true symbiotic relation with a poem, it enriches our understanding of the poet's technique and meaning.' See also Greenfield, Stanley B., 'The Canons of Old English Criticism', ELH xxxiv (1967), 141–55CrossRefGoogle Scholar. He argues (pp. 143–4) that 'even if a poem like Beowulf were to be convincingly demonstrated as of oral composition, which it has not been, the case for abandoning standard critical techniques in analyses of its poetic values remains unproved'. Greenfield then analyses some passages from Beowulf and concludes (pp. 154–5) that 'an understanding of the special technique of [formulaic] poetry rather helps the critic, as it enabled the Anglo-Saxon auditor, to evaluate the effectiveness of individual instances. We need not reject 'the ordinary canons of literary judgment' in dealing with the poetry, nor need we evolve new critical methods apart from those applicable to all English poetry.' Mutatis mutandis, the same is true of Homer.

^{page 14 note 1} The material in the Appendices (pp. 223–75) of Lord's Singer of Tales, op. cit., (p. 7 n. 1 above) is presumably among the best and most interesting of the Yugoslav songs, but it is strikingly inferior to Homer as poetry. See also Parry, Milman, 'Whole Formulaic Verses in Greek and Southslavic Heroic Song', TAPA lxxiv (1933), 179–97Google Scholar, which compares Homeric and Yugoslav formulas which have the same 'essential idea'. Parry is trying to demonstrate the similarity of the two oral traditions, but in almost every case the Homeric version is more concrete, more colourful, in short more poetic, than the Yugoslav. For example:

(49 times)

And addressing him he spoke winged words

This in Southslavic is:

Pak mu poče tiho govoriti (ii. 44, 50)

And quietly she began to speak to him.' (p. 185)

Or:

(9 times)

He spoke, and brandished and hurled his long-shadowing spear

To mu reče, bojno koplje pušti (ii. 43, 617) So he spoke to him, and hurled his battle spear.'

(pp. 187–8)

Or:

(21 times)

When appeared the early-born rosy-fingered dawn

Kad u jutru jutro osvanulo (ii. 5, 54)

When on the morn the morning dawned.' (p. 188)

Cf.Bowra, C. M., 'The Comparative Study of Homer', AJA liv (1950), 192Google Scholar: 'It would be idle to pretend that any other oral heroic poem has [Homer's] range or strength or purity of poetry…. There are fine moments… in some of the Jugoslav lays, but they are not Homeric either in spirit or in accomplishment…. The art of oral composition is still mysterious, but so are the workings of creative genius, and in the last resort not even comparative study can tell us why the Homeric poems are as good as they are.'

^{page 15 note 1} See Parry, Adam, 'The Language of Achilles', TAPA lxxxvii (1956), 1–7Google Scholar.

^{page 15 note 2} Parry, Milman, 'Studies in the Epic Technique of Oral Verse-Making. I. Homer and Homeric Style', HSCP xli (1930), 76Google Scholar: 'We must go back to the principle of Arist-archus of getting "the solution from the text", but we must enlarge it until it covers not only the meaning of a verse or passage but the poems entire, and lets us know why the poet, or poets, of the Iliad and Odyssey made them as they are…. Whatever feature of poetic art we may study, we must follow it throughout the traditional text, and try to see it clearly and fully;… [and] after finding all the elements of the poems which bear upon that feature, to draw from them when we can, but from them only, a new idea of poetic artistry….What I wish to point out is not the need of a new method, but of a stricter use, in the supreme problem of Homer's idea of style and poetic form, of the one good one.'