Sticks and Stones, and
Rhemata Hapax Legomena

by   D. E. Bynum

Of the many open questions about the corpus of ancient Greek poetry that is attributed to Hesiod, there are none more intriguing than those that pertain to the wonderful shard-pile of broken verses and narrative allusions which is all that remains of Hesiod's Catalogue of Women (Υυναικῶν Κατάλογος) and the Ehoiai ([Μεγάλαι] ᾿Ηοῖαι).1 Only four of the hundreds of fragments in the heap actually have the incipit ἠ᾿ οἵη (or like her...), and of those four, only one is of more than three lines' length. That one is of course the so-called Ἀσπίς (Herculis Scutum).2 Its subject, the treatment of that subject, features of its language, and its great length (480 verses), all argue against the poem's authenticity as a whole, and no more than some of the first fifty-six of its hexameters are usually accepted as even possibly of true Hesiodic origin (whomever or whatever that may mean).

    The first lines of the Ἀσπίς purport to be about Alkmene, the mother of Herakles; but in fact they more concern her husband, Amphitryon, by whose name Herakles was known patronymically throughout his mortal lifetime (although in reality it was Zeus who had sired him).

    Like many another displaced man of early Greek heroic legend, Amphitryon was obliged to leave his native land because he had killed someone. He thus belonged to a common type, that of the refugee homicide, like Patroklos in the Iliad (Ψ 81 ff), or in the Odyssey Theoklymenos (o 222 ff), and even Odysseus himself as he lies to Athena in Book Thirteen. Amphitryon belongs to this same type, but he is a unique multiform of it in one respect: his wife Alkmene willingly accompanied him into his exile, despite the fact that it was her own father whom her husband had killed. Hesiod (or whoever it was who composed these lines in the Hesiodic corpus) was presumably cataloguing her among the many other mortal women whom Zeus had sometime loved when he sketched how Alkmene went with Amphitryon to his place of refuge. Thus Alkmene too, no less than her husband, belonged to a common type (i.e., a paramour of Zeus), though she had a unique destiny within her type:

Or such as she, who having left the house and the land
        of her paternity
Went to Thebes with the warlike man Amphitryon,
Namely Alkmene, the daughter of Elektryon, inciter of the people.
She exceeded all of gentle womankind
5   In shapeliness and stature. Nor indeed did anyone
        rival the quality of her mind,
Not any one at all of those whom mortal men begat in
        the marriage-bed upon mortal women.
From her face and from her dark eyes
A loveliness streamed forth like unto that of golden Aphrodite.
And she held her husband in such high regard in her own heart
10   As none of gentle womankind had ever done before.
For verily, having overpowered him by main strength, he had
        killed her goodly father,
Being angry about cattle. And so he left his native country
And came among the Kadmeian shieldmen.
There he dwelt with his estimable wife
15   Without the delights of love, for it was forbidden him
To share the marriage-bed with Elektryon's trim-ankled daughter
Until such time as he would take vengeance for the great-hearted
Of his wife and with ferocious fire burn down the villages
Of the lordly Taphians and Teleboans.
20   For thus it was enjoined upon him that he should do,
        and the gods were witnesses of it.
Indeed he dreaded their wrath, and strove that he might
        as quickly as possible
Accomplish the great deed that was laid upon him
        by ordinance from Zeus.
And so together with him, longing for war and the noise of battle,
Went the Boiotians who drive horses, puffing over their
        hide-covered shields,
25   And also the Lokrians who fight at close quarters,
       and the great-hearted Phokians.
The brave sons of Alkaeus led them,
And he rejoiced in his troops. But all the while the father
        of men and of gods
Was contriving in his mind yet another scheme whereby,
        both for the benefit of the gods
And for men who traffic for gain, he might raise up
        a protector against calamity.
30   Thus he was plotting a strategem deep in his mind as he
        bestirred himself from Olympos,
Craving the coital embrace of a handsomely girdled woman
In the night-time. He came swiftly to Mount Typhaon, and
        from there
The wise counsellor Zeus passed onward to the uppermost
        Peak of the Sphinx.
There he sat himself down and meditated godly deeds.
35a|   Then, when it was night, verily he bedded
36a   and mingled in a coital embrace
35b|   With the slim-ankled daughter of Elektryon,
36b   and so slaked his longing.
And on the same night the splendid warrior Amphitryon,
        inciter of the people,
Came home from the great task which he had completed.
39   He did not however go about rousing his retainers
40a   and his herdsmen in the countryside
40b   Until he had gone in to bed his wife,
For such was the yearning in his heart which gripped
        the shepherd of the people.
46   So he lay all night long with his estimable wife
Enjoying the gifts of golden Aphrodite.
But because she had been subject both to the god
        and to a very fine man
She gave birth to twin children there in Thebes of the seven gates;
50   Yet they were not like-minded, albeit they were siblings.
One of them was inferior, but the other was a greatly better man
-- He of the Herculean strength, terrible and overpowering,
Whom she conceived while in the embrace of the cloud-wrapped
        son of Kronos;
But the spearman Amphitryon begot Iphikles.
55   Thus she gave birth simultaneously to different progeny,
        the one sired by a mortal man,
But the other one by Kronian Zeus, begetter of all the gods.
    There is not much more about Amphitryon elsewhere in the surviving classical literature.3 Homer does not rehearse the story about him, although he does know Amphitryon's name as the patronymic of Herakles (E 392), and in a Catalogue of Women of his own he includes Alkmene as the mother of Herakles by Zeus (Ξ 323-4). All else that has come down to us about Amphitryon is Alexandrian or later, and is full of specious rationalizations remote from the ancient tradition. Perhaps even in the early tradition Amphitryon was never more than a subsidiary figure in the legend of his very famous foster-son. Slight and obscure as it is in its surviving form, the story of Elekryon's and Amphitryon's dealings with each other and with the Taphians is nonetheless certainly just such matter as should have been part of an Argive or Theban "cycle" of ancient Greek epic, if ever there were such tales sung in places of Greek speech about things other than the war at Troy and its contingencies.

    How could we recognize a fragment of such extra-Homeric epos if we were to see one? Surely there would be formulas in it, no matter how variously one might choose to regard the significance of that fact. Such formulas would of course be the inherited stock of the immediate tradition that gave rise to the text concerned (whose-ever or whatever tradition it might be); but there might also be found in it some "common" formulas -- phrases which that tradition (however it were defined) would have in common with, for example, Homer's stories of Troy. Experimentally, and merely to demonstrate the principle of the thing, one might even compare the diction of Hesiod's own little tale of Amphitryon both with the rest of "Hesiod" and with Homer.

    For, short as it is, Hesiod's little text does happen to show a certain recurrency of phrase, and even recurrent phrases of several different types, within its own confines, e.g., προλιποῦσα../λιπὼν..πατρίδα γαῖαν (vv. 1 ≃ 12); πολυχρύσου Ἀφροδίτης (vv. 8 ≃ 47); ἧς ἀλόχου (vv. 18 ≃ 40). In diagramming the passage below, I have first used underscoring to mark all the phrases that are found again elsewhere in Hesiod.4 Having identified them, I sought the residue of apparent Hesiodic rhemata hapax legomena first in the Iliad5 (colored orange) and then in the Odyssey and minor Homerica (colored blue).6 The diagram thus shows: first, what is recurrent diction in Hesiod; and if not in Hesiod, then in the Iliad; and if not in Hesiod nor in the Iliad, then in the Odyssey and lesser Homerica. Where recurrencies are contiguous or overlap, the line is repeated in parentheses with the appropriate further marking.

  1    Ἢ οἳη προλιποῦσα] δόμους καὶ [πατρίδα γαῖαν

  2    ἤλυθεν ἐς Θήβας μετ᾽ ἀρήιον ᾿Αμφιτρύωνα

  (2    ἤλυθεν ἐς Θήβας μετ᾽ ἀρήιον ᾿Αμφιτρύωνα)

  3    Ἀλκμήνη, θυγάτηρ λαοσσόου Ἠλeκτρύωνος·    (cf. Σ 399)

  4    ἥ ῥα γυναικῶν φῦλον ἐκαίνυτο θηλυτεράων

  5    εἴδεΐ τε μεγέθει τε· νόον γε μὲν οὔ τις ἔριζε

  (5    εἴδεΐ τε μεγέθει τε· νόον γε μὲν οὔ τις ἔριζε)

  6    τάων, ἁς θνηταὶ θνητοῖς τέκον εὐνηθεῖσαι.

  7    τῆς καὶ ἀπὸ] κρῆθεν [βλεφάρων t᾽ ἄπο κυνεάων

  8    τοῖον ἄηθ᾿ οἷόν τε πολυχρύσου Ἀφροδίτης.

  9    ἣ δὲ] καὶ ὣς [κατὰ θυμὸν ἐὸν ἀκοίτην,

10    ὠς] οὔ πώ τις [ἔτισε γυναικῶν θηλυτεράων·

11    ἦ μέν οἱ πατέρ᾿ ἐσθλὸν ἀπέκτανε ἶφι δαμάσσας,

12    χωσάμενος περὶ βουσί· λιπὼν δ᾿ ὅ γε πατρίδα γαῖαν

12    (χωσάμενος περὶ βουσί· λιπὼν] δ᾿ ὅ γε [πατρίδα γαῖαν)

13    ἐς Θήβας ἱκέτευσε φερεσσακέας Καδμείους.

14    ἔνθ᾿ ὅ γε δώματ᾿ ἔναιε σὺν αἰδοίῃ παρακοίτι

15    νόσφιν ἄτερ φιλότητος ἐφιμέρου, οὐδέ οἱ ἦεν

(15    νόσφιν ἄτερ φιλότητος ἐφιμέρου, οὐδέ οἱ ἦεν)

16    πρὶν λεχέων ἐπιβῆναι ἐυσφύρου Ἠλεκτρυώνης,

17    πρὶν γε φόνον τείσαιτο κασιγνήτων μεγαθύμων
                (cf. ω 434, 484)

18    ἧς ἀλόχου, μαλερῷ δὲ καταφλέξαι πυρὶ κώμας

18    (ἧς ἀλόχου, μαλερῷ] δὲ καταφλέξαι [πυρὶ κώμας)

19    ἀνδρῶν ἡρώων Ταφίων ἰδὲ Τηλεβοάων.

20    τὼς γάρ διέκειτο, θεοὶ] δ᾿ ἐπὶ [μάρτυροι ἦσαν·

21    τῶν ὅ γ᾿ ὀπίζετο μῆνιν, ἐπείγετο δ᾿ ὅττι τάχιστα

22    ἐκτελέσαι μέγα ἔργον, ὅἱ Διόθεν] θέμις [ἦεν.

23    τῷ δ᾿ ἅμα ἱέμενοι πολέμοιό τε φιλόπιδός τε

24    Βοιωτοὶ πλήξιποι, ὑπὲρ σακέων πνείοντες,

25    Λοκροί τ᾿ ἀγχέμαχοι καὶ Φωκῆες μεγάθυμοι

26    ἕσποντ᾿· ἦρχε δὲ τοῖσιν ἐὺς πάις Ἀλκαίοιο

27    κυδιόων λαοῖσι. πατὴρ δ᾿ ἀνδρῶν τε θεῶν τε

28    ἄλλην μῆτιν ὕφαινε μετὰ φρεσίν, ὥς ῥα θεοῖσιν

29    ἀνδράσι τ᾿ ἀλφηστῇσιν ἀρῆς ἀλκτῆρα φυτεύσαι.

(29    ἀνδράσι τ᾿ ἀλφηστῇσιν ἀρῆς ἀλκτῆρα φυτεύσαι).

30    ὦρτο δ᾿ ἀπ᾿ Οὐλύμποιο δόλον φρεσὶ βυσσοδομεύων,

30    (ὦρτο δ᾿ ἀπ᾿ Οὐλύμποιο δόλον φρεσὶ βυσσοδομεύων,)

31    ἱμείρων φιλότητος ἐυζώνοιο γυναικός,

32    ἐννύχιος· τάχα δ᾿ ἷξε Τυφαόνιον· τόθεν αὖτις

33    Φίκιον ἀκρότατον προσεβήσατο μητίετα Ζεύς.

34    ἔνθα καθεζόμενος φρεσὶ μήδετο θέσκελα ἔργα·

34    (ἔνθα καθεζόμενος φρεσὶ μήδετο] θέσκελα [ἔργα

35    αὐτῇ μὲν γὰρ νυκτὶ τανυσφύρου Ἠλεκτρυώνης

36    εὐνῇ καὶ φιλότητι μίγε, τέλεσεν δ᾿ ἄρ᾿ ἐέλδωρ·

37    αὐτῇ δ᾿ ᾿Αμφιτρύων λαοσσόος, ἀγλαὸς ἥρως,

37    (αὐτῇ δ᾿ ᾿Αμφιτρύων λαοσσόος, ἀγλαὸς ἥρως),

38    ἐκτελέσας μέγα ἔργον ἀφίκετο ὅνδε δόμονδε.

38    (ἐκτελέσας μέγα ἔργον ἀφίκετο ὅνδε δόμονδε).

39    οὐδ᾿ ὅ γe πρὶν δμῶας καὶ ποιμένας ἀγροιώτας

40    ὦρτ᾿ ἰέναι, πρίν γ᾿ ἧς ἀλόχου ἐπιβήμεναι εὐνῆς·

41    τοῖος γὰρ κραδίην πόθος αἴνυτο ποιμένα λαῶν.


46    παννύχιος δ᾿ ἄρ᾿ ἔλεκτο σὺν αἰδοίῃ παρακοίτι

47    τερπόμενος δώροισι πολυχρύσου Ἀφροδίτης.

48    ἣ δὲ θεῷ] δμηθεῖσα [καὶ ἀνέρι πολλὸν ἀρίστῳ

49    Θήβῃ] ἐν [ἑπταπύλῷ διδυμάονε γείνατο παῖδε

50    οὐκέθ᾿ ὁμὰ φρονέοντε· κασιγνήτω] γε μὲν [ἤστην·

51    τὸν μὲν χειρότερον, τὸν δ᾿ αὖ μέγ᾿ ἀμείνονα φῶτα,

52    δεινόν τε κρατερόν τε, βίην Ἡρακληείην,

53    τὸν μὲν ὑποδμηθεῖσα κελαινεφέι Κρονίωνι,

54    αὐτὰρ Ἰφικλῆa δορυσσόῳ Ἀμφιτρύωνι.

55    κεκριμένην γενεήν, τὸν μὲν βροτῷ ἀνδρὶ μιγεῖσα,

56    τὸν δὲ Διὶ Κρονίωνι, θεῶν σημάντορι πάντων.

    Most of the phrasing marked in this passage as occurring elsewhere in Hesiod or in Homer recurs as ipsissima verba. The few instances of merely formulaic diction -- where only a type of phrasing recurs, but not in the same words -- are shown in italics. There are a good many discontinuous formulas in the passage (i.e., diction which occurs just so somewhere else, but which is interspersed with other, additional words in this passage). I have marked the latter with exclusionary brackets to isolate the interspersions (xxxx]xxx[xxxx).

    The fifty-two verses of this passage from the Ἀσπίς contain altogether about 810 syllables. Although the entire corpus of Hesiod -- both whole works and fragments -- is scarcely 3,000 lines (after excluding the Ἀσπίς beyond v. 56), it nevertheless shows a repetition of diction amounting to approximately 270 syllables in our passage, or about 33% of it.

    Various phrases amounting to approximately 218 syllables which are dispersed throughout the fifty-two hexameters cannot be shown ever to have been repeated by Hesiod, but they are found in Homer. Thus, interestingly enough, almost as much (27%) of the diction in the Hesiodic tale of Amphitryon is found recurrent in Homer as is found in Hesiod himself (33%), and this despite the fact that only Hesiod and not Homer tells the story of Amphitryon.

    Finally, the sum of the diction that recurs either in Hesiod or in Homer, or in both together, is 60% of the passage.

    But what of the remaining 40% of these verses, which seemingly consists of hapax legomena? (Or, since one is not dealing here with individual words, but rather with characteristic combinations of words, they might perhaps be more exactly called rhemata hapax legomena.) Does that nearly half of the passage which cannot be found to have been used in a formulary manner elsewhere either in Hesiod or in Homer formidably obstruct a vision of this text as being itself possibly a fragment of the ancient oral epic tradition? Yet many of those very phrases in the first lines of the Ἀσπίς whose spoors we cannot find in other traditional texts are on their faces highly epic: Βοιωτοὶ πλήξιποι, Λοκροί τ᾿ ἀγχέμαχοι, ἱκέτευσε φερεσσακέας Καδμείους, and even the whole line (un-Homeric to be sure, but no less epic for that) τῷ δ᾿ ἅμα ἱέμενοι πολέμοιό τε φιλόπιδός τε.

    Or is it rather, perhaps, that we have in such lines as τάων, ἁς θνηταὶ θνητοῖς τέκον εὐνηθεῖσαι not the debris from a lost "cycle" of ancient Greek oral epic tradition, but instead an authentic instance of that veritable basilisk (or is it unicorn?) of the medievalists, a skillfully turned literary phrase in a "transitional text" that only emulates the oral tradition which preceded it?

    If only we possessed some other, more ample, and more probably oral traditional epic texts of the story of Amphitryon and others like it! Then surely we would be able to determine not merely as a matter of greater or lesser hypothetical possibility, but quite simply as a matter of fact how much of its language might be formulary, and what formulas in particular were characteristic of that story in the oral tradition. Indeed it would be helpful for better understanding how not only the Hesiodic poetry was made but also how the Homeric poetry was made could we but consult the rest of the seventy epics that were reputed to have been in the Alexandrian Library. Or, failing that, one could no doubt make good use of the full texts of only a judicious selection from among them. Not to seem greedy, I personally could content myself with only the Minyas, Kypria, the original Thebaid, the Epigonoi, Naupaktia, and Iliu Persis.

    But wishes are not horses, and we must come at the present problem realistically. For no such other epics have survived from ancient Greece. And for every common formula which we are able to identify as between Homer and other, more fragmentary ancient poetry, a score of apparent rhemata hapax legomena rise in the fragments to call in question whether what seem to be common formulas really are such, or whether they are on the contrary only literary borrowings from, or literary imitations of, an oral tradition.

    The solution to that problem, as of many another concerning early poetry, awaits specific further developments in general understanding about the nature and uses of formulary language in oral traditional poetry. Nor is there any way at all to arrive at the necessary improvements on the science of formulary diction only within the confines of the surviving ancient Greek poetry, which is simply too scantily preserved for that purpose. A process of analogical deduction from the evidence of some more fully preserved tradition of a comparable kind in some other language is thus not merely a possible alternative or a supplementary adjunctive to direct examination of the early poetry; it is the only passable avenue to the desired end.

    And even apart from questions about Homer and Hesiod, if it is to South Slavic oral epic traditions that one looks for analogue, the findings there may also have considerable explanatory power about the peculiar poetic accomplishments of the Slavic-speaking peoples in the Balkans, which are interesting in their own right. For in their manner of living, where the force of tradition was in general very great, and where tradition mostly excluded the making of poetry or the cultivation of poetic experience by writing in native languages, the native traditions of singing poetry without writing continued until very late in the modern age to produce more good poetry of more profound and enduring interest to more of the population than happened anywhere else in Europe so far as the historical record shows. How this happened is intrinsically interesting.

    Hesiod's story of Amphitryon has a counterpart in modern South Slavic oral epic tradition, and a counterpart which is moreover much nearer to Hesiod's own story than anything surviving from ancient Greek tradition. Avdo Međedović, the best singer in the South Slavic tradition who happens to have become known outside it, told the story to Milman Parry twice in 1935; once in song, and once by dictation. Both the texts of his telling run to more than a hundred times the length of the fragment in the Ἀσπίς (which, needless to say, Avdo neither knew nor could have known).7

    Amphitryon's name has been variously interpreted as meaning "much tried; much worn (experienced)"; or alternatively, "he of the wide dominion." The former sense has more often been preferred. It is both defensible on linguistic grounds and obvious enough in the formal contruction of the word to be a good folk-etymology as well. Avdo's corresponding hero is Ali Vlahinjić. He is a man not only without a patrimony (as was Amphitryon during his exile8) but also without a patronymic, a most unusual social deficiency in the South Slavic tradition. Thus, his last name, Vlahinjić, makes him solely his mother's son. Like Amphitryon, he is so young that Avdo's story about him concerns his first courtship and marriage; but again like Amphitryon, he is already an extraordinarily seasoned warrior despite his youth. Vlahinjić has been his emperor's champion in no fewer than thirteen ritual duels before the events in Avdo's story even begin. Other notable heroes of this tradition may live and fight valiantly into old age without ever accomplishing such a number of distinguished labors.

    The well-born woman whom Vlahinjić wishes to marry pledges herself to him against all custom and every counsel of common sense. He is, in the eyes of others, the very worst choice she could make among a host of more seemly suitors. Yet, like Hesiod's Alkmene, she honors her chosen husband in her heart more than any other of womankind has done before her, and she determines that she will either marry him, or else no one. And again like Amphitryon, he too kills an older man among his people (precisely one of a sufficiently elder age to be either his own or his bride's father), and then forsakes the place of his nativity to live thereafter with his wife in another jurisdiction.

    Although she has dedicated herself utterly and unalterably to him against what appear to be her own better interests, he is however forbidden to consummate their marriage until such time as he will have completed a perilous military labor of vengeance against the long-unpunished murderers of her brothers. All the lords of the place are witnesses to the imposition on him of this task and his oath to accomplish it quickly. This he does by making an expedition to a distant location where he slays a prodigious number of the murderers' dependents and burns their outlying wooden dwelling with ferocious fire. That done, he returns homeward (not, that is, to his native land, but to the place of his new, adoptive domicile), and goes straight to his wife without wishing to awaken or parley with any intervening males. But just when he is about to claim her, another male who is senior to him comes surreptitiously down into the story from the heights whence he has been observing and plotting a strategem, and claims the woman for his own, so that Vlahinjić cannot bed her until she is released from the usurper's control.

    To this six-thousand-line version of the Amphitryonic tale we can now address those same questions about diction which the Hesiodic fragment and its surrounding corpus of Hesiodic hexameters could not answer, not even with the help of mighty Homer. For now we are able to examine in a single poem more than twice the number of verses which the whole surviving corpus of Hesiod can afford, while the corpus to which this one South Slavic epic belongs -- that of Avdo Međedović -- runs to more than eighty thousand lines. Here we have the necessary amplitude of evidence which is wanting in the extant ancient Greek epos, and we unequivocally know it to be, moreover, the direct, unmediated result of oral traditional composition.

    The original question which has brought us to this point concerns whether phrases of seemingly traditional poetry which happen not to be repeated in extensive bodies of such poetry argue in some degree against that poetry's origin in an oral tradition. And further: what may the relationship be in respect of diction between poems composed in an oral tradition at different times by different makers? And again, together with these two theoretically distinct but pragmatically inseparable questions there is inextricably connected a third: do oral epic traditions depend upon their singers' mental retention and repetition of specific poetic phrases or formulas (which they must presumably have heard and learned sometime from other singers before them) on the one hand, or only upon certain techniques which they have learned for making useful phrases of certain traditional sorts (regardless of whether exactly those phrases had ever occurred to them before, or might ever be used again) on the other hand.

    The poetry of Avdo Međedović aptly responds to these several questions.

    Avdo used to begin his epics (including the one about Ali Vlahinjić, from his sung version of which I draw the next examples) by saying that the narratives in them belonged to the time of the most illustrious Turkish sultans. On this occasion, he said specifically that the events of his tale happened

A u zeman sjajnog Sulejmana,
najvećega na svijet' vladara,
koji s trista šes'et valiluka
vladovaše i gospodovaše.9     13
    Now the last of these lines was very useful for stating exactly what it was that Suleyman the Magnificent did in his illustrious age: he "governed and was lord of his dominions." That was something that Avdo had occasion to say elsewhere too, not only about Suleyman but also about other prominent characters in his various tales, whose number was not small. Yet despite the seeming utility of the line, its catchy internal grammatical rhyme, and the neat paratactic balance between its two hemistichs, Avdo never used it again in more than 60,000 verses of further epic composition. He had considerable occasion in those 60,000 verses to express the same essential idea, but did not again formulate it in that same way. Instead, the line vladovaše i gospodovaše appears only once, as a phrase of a certain kind; a kind which he did however happen to use from time to time apart from that particular line. Thus, on other occasions he said, for example, concerning not the sultan but rather certain of his male subjects and their horses:10
zauzdaše i opusatiše     5771
siguraše i natovariše     9113B
Or, concerning a young man and woman:
pogledaše i begenisaše     8371A
ogledala i begenisala     3251A
sevdisala i begenisala     1921
Or again, about an older woman:
hankovala i gospodovala     13092B
But although phrases of this general kind happened repeatedly in Avdo's composition, never in more than 37,000 lines did any of the specific formulations which I have just cited happen more than once. Only in the infinitive mood was one of their general kind actually repeated once verbatim, and once again nearly so:
carovati i gospodovati     8109A, 11905A
carevati i gospodovati     7989A
    Finding such instances as these, one naturally asks oneself whether the formula at hand - carovati i gospodovati -- really is a phrase that the singer has remembered wholly and as such from his past experience of hearing other singers use it, reinforced thereafter by his own subsequent use and reuse of it; or is it only a phrase of a certain kind which he has himself happened to compose according to a pattern somehow lodged in his mind and so repeated twice during the final four thousand lines of a 12,000-line epic, but never again exactly so thereafter? And does the repetition only thrice of carovati i gospodovati mark it as genetically different from vladovaše i gospodovaše which Avdo happened not to repeat even once? Is the latter phrase necessarily any less wholly remembered than the former one merely because the one recurs once or twice in the particular singer's diction, while the other does not? In a word, what exactly is the balance -- or, should one prefer to say, the interaction -- between remembering and making poetic statements in an oral tradition?

    The problem is plainly insoluble by inspection of only a single singer's diction -- which is somewhat paradoxical, since it also cannot be solved without meticulous attention to the repetitive habits of the single singer. We cannot tell from Avdo Međedović's recorded poetry alone whether there is or is not any significant difference -- nor exactly what the difference might be -- in the genesis of the two phrases vladovaše i gospodovaše and carovati i gospodovati, but at the same time we cannot even pose the problem of whether there may be a meaningful difference between them until it has been established, as I have just done, exactly how frequent or infrequent those phrases are in a large corpus of one good singer's poetry.

    Were vladovaše i gospodovaše to have occurred as a whole-line repetition twenty times in 6,000 lines, let us say, or even twenty times in 60,000 verses of a single singer's poetry, such a frequency of exact recurrence might be strong reason for supposing that the line were an entirely remembered formula and in no sense merely made up upon some general scheme of diction to meet the metrical and ideational need of the moment ad hoc. And in fact there are plenty of such oft-recurring lines in Avdo's epics, just as there are in Homer. The present case is, however, quite different. The surpassing rarity of the expression vladovaše i gospodovaše as a whole line strongly encourages the counter-inference, namely that it was a construction ad hoc, a formulation independent of the singer's necessarily ever having heard, or remembered, or even himself in some quiet moment of private reflection in the past having composed and now recalled exactly that previously formulated phrase and no other.

    Depending upon how one chooses to regard it, the line carovati i gospodovati may also be perceived chiefly in the light of its rarity; it seems to be, as it were, accidental. Yet it retains just enough currency of repetition to suggest with equal force, if for any reason one is predisposed to feel it, that the line is possibly not only an offspring of the moment, but rather in some measure a piece of the tradition in its own right, an entity superior to the mere scheme of such-like phrases, in which Avdo was clearly very fluent.

    So the single singer's diction, taken by itself, poses a crucial problem very well, but offers in itself no means of solution. And even after examining this problem in the big corpus of the modern singer, we remain no farther advanced toward a solution than we were when dealing with Hesiod alone, although the palpable likeness of the two experiences certainly does much to clarify the nature of the problem itself, and its universality.

    But if it were to happen that a particular phrase of very great infrequency in Avdo's diction (which we know de facto to be oral traditional) had some considerable currency in the poetry of other traditional singers (as with Hesiod's rare phrases which appear in Homer), we might still be able to say with great certainty that the phrase was a remembered piece of the oral poetic tradition in its own right, current in the tradition as a common formula, and not just an extrapolation by Avdo from some general pattern or patterns of poetic phrasing to suit the metrical and ideational exigencies of the given moment in the telling of a particular epic tale. Furthermore, the most useful phrases for comparison in this manner would be those gotten from the greatest practical local distance from the poet whose own practice one is examining, since they would best show the probable currency of themselves over the greatest range beyond the neighborhood of the particular singer in question.

    For this reason I turn next to a text of oral narrative poetry in a Macedonian dialect from the district of Prilep. It was collected by M. Tsepenkov and published for the first time in the second volume of the Sbornik za narodni umotvorenija, nauka i knižnina (p. 116 ff). The singer was Ristet P'rtsan, a shoemaker (or more precisely, a maker of slippers, or papudžija) in Prilep town. The text is balladic -- only 366 verses in length -- and tells how Krali Marko (the modern Macedono-Bulgarian Herakles) once boasted of his great strength at a banquet of seventy kings. Marko told them, to wit, that if only somewhere the Earth had a handle by which to grasp it, he would lift it and turn it over. The seventy kings were scandalized by so unseemly a proposal, ridiculed Marko for such a boast, and reproached him roundly for boasting at all. They tell him that his father V'lkašin, who was a very rich man (and hence a very worshipful one) never behaved in so arch a fashion, and so neither should he.

    Marko is insulted, calls for his horse, and leaves the banquet. On the road homeward, he is met by the Lord God, who speaks to him, asking him to dismount:

Vaka mu Gospod progooril:
--Oj, junače, neznaen delija,
ti što nosiš dolgata gargija     40
ja podslezi ot konja nazemi!
Thus did the Lord speak to him:
"Hail to thee, good man, o thou unknown warrior,
thou who bearest the long spear,     40
get thee down from off thine horse!"
    The phrase neznaen delija (unknown warrior, warrior-stranger) happens again later in the same text verbatim. Marko goes to a blacksmith to have his horse shod and his sword sharpened before he undertakes an exploit. The blacksmith is afraid of him however, and replies to Marko's request:
--A oj mi si neznaen delija:     120
jaz ne možam konja da ti koam...
"Oh thou warrior, stranger to me,     120
I cannot shoe thine horse..."
The blacksmith prevails upon Marko to wait a week until he has time to do the work, and sends his daughter-in-law to serve the hero until then in a wineshop. At the end of the week, Marko pays her well for her service, but tells her father-in-law the blacksmith that he knows how envious women are. For that reason he insists upon giving a gratuity also to the blacksmith's daughter Angelina, a girl who is betrothed but not yet married. Angelina is reluctant to wait upon Marko, but finally agrees to go to him on condition that her father will permit her to wear other clothes than those which she ought more properly to wear as an affianced girl. With that understanding, says she:
--Togaj, tetko, pred nego ke odam,
pred onega neznaen delija,     187
pred onega težka pijanica!
"So, then, daddy, I shall go to him,
To that unknown warrior,     187
To that hard-drinking man!"
    When she appears before him to receive her gratuity, Marko steals her, putting her in the saddle behind himself, and flees homeward with her. Her fiancé pursues them until they reach the banks of a river, where Marko has a bad moment:
Reka bila lošo dotečena,
što nosila d'rvja i kamenje.     304
Togaj Marko beše se uplašil
ot rekata, deka beše stekla,
i ot junak, deka go pritasal.
The river was in full flood,
carrying along timbers and boulders.     304
So Marko took fright both
at the river, which was so swollen,
and because of the man who was pursuing him.
At this terrible moment, Marko's horse Šarec speaks to him and suggests a tactic:
--Dejdi, Marko, moi stopanine,
strede reka ti si udri kopje,
za da zaprit d'rvja i kamenje,     315
dori da si nie pomineme.
Ti ne boj se, jaz ke si izplivam,
jaz sum učen ušte od malečko!
Udril Marko kopje strede reka
i mi zaprel d'rvja i kamenje.     320
"Come now, my master Marko,
cast your spear into the river
and it will balk the timbers and the boulders     315
so that we may get across.
Fear not, for I shall swim the river for you readily,
since that's a thing I learned to do when I was still a colt."
So Marko hurled his spear into the river
and thereby balked the timbers and boulders.     320
    The horse is as good as its word, but Marko's pursuer also attempts to cross the river:
Svikal Šarec na Marko kraleta:
--T'rgaj, Marko, kopje otsred reka,     325
da ponesit d'rvja i kamenje,
da preprečit konja bedevija!
T'rgnal Marko kopje otsred reka
i poneslo d'rvja i kamenje,
ta preprečil konja bedevija.     330
I se vratil junak ot Krajina,
ta zastanal Marka da go gledat.
Šarec shouted to King Marko:
"Now, Marko, wrench your spear out of the river     325
and let it tumble its burden of timbers and boulders
to stop [our pursuer's] bedouin mare!"
Marko wrenched his spear out of the river's flood
and onward once again it tumbled its burden
        of timbers and boulders,
which stopped the bedouin mare.     330
So the man of the Border turned back from his pursuit
and stood gazing at Marko from the far bank.
    Marko is so grateful to his horse Šarec that he kisses it between its eyes. Seeing this, the cheated bridegroom calls to Marko across the river:
Dejdi, more, neznaen delija!     340
Ne celivaj konja šarenoga,
tuk celivaj gjuzel Angelina.
"Come now, foolish fellow, thou unknown warrior,
kiss not thy piebald horse;
kiss instead the lovely Angelina."
    So in the course of this 366-line ballad Marko meets four persons: the Lord God, the blacksmith, the blacksmith's daughter Angelina, and her financé, the otherwise nameless junak ot Krajina (man of the Border). All four speak to Marko, addressing him each time as neznaen delija (unknown warrior, warrior-stranger). That phrase is a classic noun-epithet formula. Its place in the South Slavic decasyllable is regularly in the second hemistich. Every time the singer P'rtsan had a character outside Marko's usual social circle address him, the speech is introduced by this formula. Even the horse Šarec, when it speaks to Marko for the first time at Marko's acquisition of it earlier in the poem, calls Marko neznaen delija:
--Oj junače, neznaen delija,
ja kaži mi čie si koleno,
jaz da sum ti tvoja verna sluga.     119
O thou man, unknown warrior,
Tell me of what family thou art sprung,
that I may be thy faithful servant.     119
    The Serbo-Croatian singer Avdo Međedović also sang an epic about a youthful hero whose special horse swam a flooding river to steal a bride. While in that bride's native country, the hero drank wine prodigiously in an inn where a youthful but sisterly woman served him. Through her he met his own foreign bride-to-be, whom he then stole. Having taken the bride, he too was pursued as far as the river, which he was able to cross, but not his pursuers. The hero of this epic was the same South Slavic Amphitryon, Ali Vlahinjić, already described above. Avdo first sang and then dictated two complete texts of the epic, each about 6,000 lines long. Yet once and only once, in the sung telling of the story, the wine-maid speaks to the hero and says:
šta ti kazah, neznana delijo     3858
I have already told thee my decision, unknown warrior!     3858
    The phrase in the second hemistich, neznana delijo, is the equivalent in Avdo's dialect of Ristet P'rstan's neznaen delija. Avdo did not, however, use it as P'rtsan did, five times in 366 verses. Instead, he used it only once in 6,048 verses. He subsequently retold the whole epic with the same events in it, but without again using this particular noun-epithet formula even once. The two texts together, the sung and the dictated, amount to nearly 12,000 lines. If we look back to the 12,000 verses which Avdo happened to compose during the same month in 1935 immediately before he made the two texts of the one epic of Ali Vlahinjić (for those earlier verses happen to have been recorded also), not a single instance -- not even a paraphrase or a grammatical variant -- of neznana delijo can be found in them.

    And similarly, if one looks also to the next 12,000 verses of text that Avdo made after completing the two tellings of the Vlahinjić epic, one may thus extend the field of search to more than 36,000 lines and still find only the one unique occurrence of this formula. To put the case statistically, the formula is more than five thousand times rarer in the given sample of Avdo's poetry than in P'rtsan's, although the story they both tell is the same. An important lesson proceeds from such evidence: the frequency of a formula's recurrence in any given singer's compositions is unrelated to its status as a common formula in the oral tradition to which that singer belongs.

    The same principle applies to other types of formula besides the noun-epithet kind. A very similar relationship exists also between P'rtsan's paratactic formula of the second hemistich d'rvje i kamenje (timbers and boulders) and its counterpart in Avdo's diction. It too is a common formula in the South Slavic traditions of oral narrative poetry. In the same passages which I have already quoted above from P'rtsan's poetry, he used that phrase and only it to describe repeatedly the peril of the flooding river. In the nature of the story, that phrase had no other utility than to describe the river, hence it did not occur outside that section of his narrative. But within that section, it was intensely useful to P'rtsan, who used it four times in a mere twenty-four lines (vv. 244-266). As in the case of the phrase neznana delijo, Avdo however used his counterpart of d'rvje i kamenje with exponentially less frequency, employing it only once in hundreds of lines devoted to the description of Ali Vlahinjić's raging river:

Ona drvlje i kamenje valja     2543
It [the river] tumbled timbers and boulders     2543
Avdo's highly developed habit of inverting word-order does not conceal the fact that drvlje i kamenje belongs to the second hemistich in his composition no less than does its counterpart in P'rtsan's verses.

    A Serbo-Croatian text which the Serbian collector Vuk Karadžić obtained by dictation from a certain haiduk named Stojan in the early nineteenth century concerns the paternity of the South Slavic Hercules, Marko Kraljević (Mark the King's Son). The king whose son he was according to this 305-line ballad was a certain Vukašin by name, who coveted another man's wife. In order to take her for his own, he had to reach the city where she dwelt. Beneath its walls there flowed a dangerous river:11

Mutna teče Tara valovita,
ona valja drvlje i kamenje     18
The murky Tara river flowed there, full of waves,
Tumbling timbers and boulders     18
    That the phrase drvlje i kamenje is a common formula in South Slavic we already knew from the previously cited examples in Macedonian tradition and in the poetry of Avdo Međedović. The example from Karadžić's singer Stojan tells us further that the formula was already common in South Slavic at least a century before Avdo used it. But it tells us more than that; namely, that as a common formula drvlje i kamenje belonged to the description of a perilous river specifically in a story of bride-stealing. Although it was not a frequent formula in South Slavic, when this common formula did occur, that is the kind of story where it "belonged." Exactly how old this formula and the story which occasions it were in South Slavic cannot be determined, since there are no texts dating from the earlier centuries of Slavic presence south of the Danube.

    But the story itself and a paratactic formula in it describing the tumbled timbers and boulders in its dangerous river are all much older than the Slavs in regions south of the Danube, for there is a great and famous story of (multiple) bride-theft with just such a river in Homer. Twice in the Iliad the river Skamandros joins its waters with those of other streams on the Trojan plain to make a flood full of waves that tumble timbers and boulders. In these two places (but nowhere else again in either Homer or Hesiod) the timbers and boulders are a formula occupying the first colon of the second hexameter in a distich of uniform scansion:12

Μ27    αὐτὸς δ᾿ ἐννοσίγαιος ἔχων χείρεσσι τρίαιναν
Μ28    ἡγεῖτ᾿, ἐκ ἄρα πάντα θεμείλια κύμασι πέμπε
Μ29    φιτρῶν καὶ λάων, τὰ θέσαν μογέοντες Ἀχαιοί
And holding the trident in his hands the Earthshaker himself
guided [the waters] and with their waves routed out all the foundations
of timbers and boulders which the toiling Achaians had put there.
Φ313    ἵστη δὲ μέγα κῦμα, πολὺν δ᾿ ὀρυμαγδὸν ὄρῑνε
Φ314    φιτρῶν καὶ λάων, ἵνα παύσομεν ἄγριον ἄνδρα
and he [Skamandros] cast up a great wave and made a mighty commotion
of timbers and boulders, in order to stop the uncouth man


Μ28, Φ313   ˗ ˗ ˗ ˇ ˇ  ˗ ˇ ˇ ˗ ˇ ˇ ˗ ˇ ˇ ˗ ˇ
Μ29, Φ314   ˗ ˗ ˗ ˗ ˗, ˇ ˇ ˗ ˇ ˇ ˗ ˗ ˇ ˇ ˗ ˇ

It is accordingly highly probable that Homer's phrase φιτρῶν καὶ λάων was a common formula in the ancient Greek oral epic tradition just as drvlje i kamenje was in the modern South Slavic tradition.

    Both the noun-epithet formula neznana delija/o and the paratactic pair drvlje i kamenje occur only once each in a corpus of Avdo Međedović's poetry larger than that of Homer. A single short Macedonian text is sufficient to show, however, that the rarity of those phrases in no way indicates their invention ad hoc by Avdo. They are not only proper formulas, but also common formulas, i.e., formulas known and used by many other singers widely dispersed in the South Slavic tradition(s) of oral narrative poetry. They as single hemistichs were just as current, and current for the same reasons, as the entire tale of the hero who conversed with the horse that crossed the swollen river to foil his master's foes during an episode of bride-theft.

    P'rtsan's use of the two common formulas which we have just been considering was to repeat them rather often in one and the same tale. They stand out in his poetry both because his poems are short, and yet full of such iterations; that is what most of all suggests that they should be sought in other poets' diction also. Others of his lines make the same suggestion, for example:

da prepričit konja bedevija     326
ta prepričil konja bedevija     330
The formula in the second hemistich of these lines is common to the diction of other South Slavic singers in Macedonian, Bulgarian, and Serbo-Croatian alike, but it happens not to occur in Avdo's epics. To him the Serbo-Croatian equivalent of this phrase must have been something of a contradiciton in terms, since in his usage konj (horse) was masculine by natural as well as by grammatical gender, whereas bedevija (Bedouin) was to him specifically a bedouin mare (as it indeed usually was to the Bedouin themselves). Not every line or hemistich which the lesser singer P'rtsan repeated was therefore necessarily also a formula used in common with every other singer, even though it might still legitimately be called a "common formula." Accordingly, the communality of formulas does not mean that every singer in the tradition necessarily used every such formula.

    Another pair of lines in P'rtsan's poem has a construction quite similar to that of his verses 326 and 330; they refer to the counterpart of the pursuing fiancé'skonj bedevija (bedouin [female] horse), namely Krali Marko's stallion Šarec:

ne tselivaj konja šarenoga     340
jaz tselivam konja šarenoga     346
kiss not the piebald horse     340
I kiss the piebald horse     346
The second hemistich of these lines (konja šarenoga -- piebald horse) was also a common formula in the South Slavic tradition as a whole, where it was used to designate many a horse besides Marko's famous Šarec or Šarac; yet it too occurs only once in more than 37,500 verses of Avdo Međedović's poetry:
na njegova konja šarenoga     5026B
on his piebald horse     5026B
This is a fine example of the relationship among formulas of a given kind which I have elsewhere described as familiation, and it shows that not only individual formulas but even whole families of them may together be common to an entire tradition of many singers.

    Occasionally a phrase which is rhema hapax legomenon in P'rtsan's little tale of 366 verses was nevertheless a common formula used also by Avdo in his long epics. Thus the noun and its epithet in the second hemistich of P'rtsan's v. 360:

pa mi vikna stara egumana
and he called the old abbot
This formula happened once in Avdo's dictated epic,
Ili oba stara igumana     5292d
or both the old abbots     5292d
and again with a paronomastic Muslim slur in his longest sung epic:
i četiri luđumana stara     8385B
and four crazy old abbots     8385B

    Another common formula of noun and epithet in P'rtsan's poem is in the second hemistich of his verse 220:

begat Šarec kako sivi sokol
Šarec galloped swiftly as a grey falcon (on the wing)
P'rtsan did not repeat the phrase, but Avdo did:
Aljo s konja kao soko sivi     5465d
Momak skoči kā i soko sivi     4904A
Ali alit from his mount like a grey falcon     5465d
the youth leapt like a grey falcon     4904A
It made no difference to the communality of this formula that different singers understood its meaning differently. P'rtsan (and others) might use it as a simile for a horse, but to Avdo it was applicable only to the description of men. Shorn of its two-syllable conjunction, this same noun-epithet phrase served Avdo as a metaphor for a man also in his first hemistichs, in which position he indeed produced it more frequently:
sivi soko ne o'sjede dora     3527d
sivi soko ruke na silaha     1850d
sivi soko u podrume dođe     5287d
sivi soko Pavičević Luka     9385B

    The utility of another of P'rtsan's common formulas was in description of visit to a large city:

Vtasal junak vo Soluna grada,
šetkal, prašal niz tesni sokatsi     99
The hero went out upon the town of Salonica,
strolling and raising dust along the narrow streets     99
Avdo knew the formula and used it the same way, to describe the progress of pedestrians in a city:
dok begleri u sokak uljegli
u Stambolu, u sokake tesne     11514
until the beys came to an avenue
in Istanbul, and entered the narrow streets     11514

    The common formulas found as between P'rtsan's and Avdo's telling of the Amphitryonic tale in South Slavic are not discovered there merely by happy coincidence. The tale does show certain formulary constants which are characteristic of it no matter what singer tells it; but the tale need not be the same to share other common formulas of wider utility. To demonstrate this reality, I have chosen entirely at random yet another Macedonian text, but this time one that does not follow the narrative pattern seen in Hesiod's Ἀσπίς. It is number 143 in the Miladinov Collection13 and it tells how a black (i.e., a Maghrebian) Arab defeated and took captive six famous heroes: Krali Marko, Tatomir, Debel Novak, Grujica, Jankula Vojvoda, and Milo Ogrijanin. Finally a seventh hero, Childe Sekula (Sekula Detence), meets the Maghrebi, defeats him, and frees the six captives.

    The text is 442 verses long. In it, as in P'rtsan's song, the southern South Slavic singer's habit of reiterating some of his formulas encourages an expectation borne of experience that precisely they will be common to the tradition at large. So, for example, in this text it is said of a hero mounting his horse,

Se pref'rli konju na ramena     267
He flung himself onto the horse's shoulders
and again:
Si se f'rli na kon na ramena     304
He flung himself onto the horse's shoulders

    Međedović was very discriminating about horses. His version of this common formula specified the color of the horse, and it was part of a whole-line formula in his diction:

sa kamena doru na ramena     953
pa s kamena doru na ramena     2447
Off the mounting block onto the bay horse's shoulders    953
Off the mounting block onto the bay horse's shoulders     2447

    When the Maghrebian Arab captured Marko in the Miladinovs' text no. 143, the singer said of him:

pišti Marko kako ljuta zmija     96
Marko hissed at him like an angry snake
The same formula was repeated in the plural each time the Arab defeated another Christian hero and added him to the company of captives:
pištat oni kako ljuti zmii     133, 159, 185, 213,
        237, 242
they hissed at him like angry snakes
Avdo also used his version of this common formula in both the singular and plural:
fakli pištu kā i guje ljute     1672A
the brands hissed like angry snakes
tanka beše kā i guja ljuta     622B
it was slim as an angry snake
The Miladinovs' singer used this formula seven times in 442 verses, but Avdo only twice in more than 37,000 lines.

    As was previously seen in P'rtsan's song from Prilep, so again in the present text: the mere fact that a southern South Slavic singer repeats a phrase in one of his relatively short songs may suggest, but never certainly predicts, that it will be found also in the diction of Međedović's long songs, not even when the phrase is otherwise obviously a common formula among other Serbo-Croatian singers. Two good examples of this from the Miladinovs' text no. 143 are in the lines

da izpadniš na junački megdan     329
ta izpadni na junački megdan     335
pa izpadna na junački megdan     337
present your-(him-)self on the field of combat     329, 335, (337)
pet tovara za ladna meana     419
da sednime na ladna meana     424
ka sednakha na ladna meana     426
five loads-worth for the cool wineshop     419
let us take our ease in the cool wineshop     424
when they had sat down in the cool wineshop     426
Although both na/za ledna meana and na junački megdan are frequent common formulas in the South Slavic tradition generally, Avdo happens not to have used them in 37,000 lines of his epics.

    Phrases that occur only once in a given southern South Slavic text, which therefore cannot decisively be shown to be formulas on that evidence alone, do however sometimes have comparatively great confirmation as common formulas in Avdo's diction. So, for example, when the black Arab of Miladinov no. 143 finally falls in single combat with Childe Sekula, the singer said only once:

on si lesna na zelena trava     350
he fell outstretched upon the green grass
With Serbo-Croatian treatment of the preposition and desinences, the second hemistich was relatively abundant in 37,000 verses of Međedović's epics:
drugi im zelen do zelene trave     3769
pa sa sapi po zelenoj travi     5795
jedan zelen do zelene trave     2905d
savrh glave do zelene trave     5751d
odvrh glave do zelene trave     12344B, 13039B
povrh glave do zelene trave     12785B
savrh vrata do zelene trave     13006B


    It has been observed in all of the foregoing instances that the statistical frequency of any particular phrase in the diction of an oral traditional epic poet can never be taken as evidence that he has himself invented it. His phrases may be common formulas regardless of whether they are rare or frequent in his compositions. He need not make a phrase often to remember it. Indeed, those very phrases which other singers may utter most, most repetitiously, and have most in common with each other, and which therefore the best oral traditional poet is most likely to have heard from others, are the very phrases which he may himself use least in enormous amounts of his own singing. It is not that he deliberately uses fewer of the tradition's set phrases or invents more of his own, but rather that he is so proficient in the construction and combination of so many common phrases that the statistical incidence of any one of them may be greatly diluted by the presence also of so many others in his diction. He deploys his phrases in the same narrative contexts and in the same grammatical constructions, and even in the same places in his lines as do other singers, but he has so many other phrases ready to mind that he repeats himself far less in the aggregate than others do.

    This ability to regenerate and to concatenate rapidly tens of thousands of phrases which not he but the whole tradition before him has wrought remains the primary characteristic of the best oral traditional epic singer. The only sure proof of what is or is not common formula must come therefore not from the best of singers, but rather from the mediocre and weaker ones by comparison with the best. To understand how the best singers compose, one must understand the lesser ones also, for their simpler dictions often display more plainly the operative principles which govern the best poet's more complex habits.


1.   Friedrich Solmsen, R. Merkelbach, and M. L. West, eds., Hesiodi Theogonia Opera et Dies Scutum (Oxford: Clarendon Press, 1970, reprinted 1983).

2.   Op. cit., pp. 86 ff.

3.   See Pauly-Wissowa, Realencyclopädie der classischen Altertumswissenschaft, v. II (1894), pp. 1967-8; W. H. Roscher, ed., Ausfürliches Lexikon der griechischen u. römischen Mythologie, v. 1, 1 (Leipzig: B. G. Teubner, 1884-6, pp. 322-23).

4.    William W. Minton, Concordance to the Hesiodic Corpus (Leiden: E. J. Brill, 1976).

5.   Joseph R. Tebben, ed., Concordantia Homerica Pars II Ilias (Hildesheim: Olms-Weidmann, 1998).

6.   Joseph R. Tebben, ed., Concordantia Homerica Pars I Odyssea (Hildesheim: Olms-Weidmann, 1994); for the minor Homerica, Henry Dunbar, A Complete Concordance to the Odyssey of Homer (edition revised and enlarged by Benedetto Marzullo) (Hildesheim: Georg Olms, 1962).

7.   The texts are in: David E. Bynum, ed., Serbo-Croatian Heroic Songs, v. 6 (Cambridge, MA: Harvard University Press, 1980), pp. 1-152.

8.   The problem of patrimony (and, indeed, of even knowing certainly the identity of one's own father) persists in Amphitryon's lineage, according to the later Greek mythographers, who make his foster-son Herakles suffer in the same way: not knowing of his paternity by Zeus on the one hand, and on the other having the kingship which Zeus intended for him usurped by Eurystheus.

9.   Op. cit., p. 73.

10.   Here and throughout, the line-numbers referring to Avdo Međedović's poems which have no alphabetic termination designate places in the sung text of the "Ženidba Vlahinjić Alije" in Bynum, op. cit. Line-numbers terminating in a lower-case d refer to the dictated version of the same epic in the same volume. Line-numbers terminating in an upper-case B refer to the text of "Osmanbeg Delibegović i Pavičević Luka," also in Bynum, op. cit. Finally, line-numbers terminating in an upper-case A refer to the text in David E. Bynum, ed., Serbo-Croatian Heroic Songs, v. 4 (Cambridge, MA: Harvard University Press, 1974).

11.   "Ženidba Kralja Vukašina," no. 24 in Vuk. St. Karadžić, Srpske Narodne Pjesme, second edition (Belgrade: Prosveta, 1958), pp. 104 ff.

12.   David B. Monro and Thomas W. Allen, eds., Homeri Opera, I-II (Oxford: Clarendon Press, 1920).

13.   M. Arnaudov, ed., Братя Миладинови, Български Народни ПѢсни (Sofija, 1942), pp. 207 ff.

   Lines 42-45 are omitted because impugned as an interpolation.

* *

Version History

* * *

Return to Documents relating to the Milman Parry Collection