The meaning of Homeric εὔχομαι through its formulas

Chapter 3. εὔχομαι in secular contexts

Analytic Table of Contents

A. γένος εὔχομαι εἶναι and its transformations (pp. 69–78)
      Classified list of attestations (p. 69):
          I: γένος εὔχομαι εἶναι
          II: γένος εὔχομαι
          ΙΙΙ: εὔχομαι εἶναι + place/father
          IV: Doublet
                    a. Rationale of classification (pp. 69–76)
                    b. Meaning of εὔχομαι in this usage (pp. 76–78)
B. εὔχομαι εἶναι + comparative/superlative (pp. 79–83)
      Classified list of attestations (p. 79):
          I: εὔχομαι εἶναι + comparative
          II: εὔχομαι εἶναι + ἄριστος
                    a. Rationale of classification (pp. 79–80)
                    b. Meaning of εὔχομαι in this usage (pp. 81–83)
C. εὔχομαι εἶναι + social relationships (pp. 83–88)
      Classified list of attestations (p. 83):
                    I: ξεῖνος, ἱκέτης
                    II: Miscellaneous
      Discussion of attestations (pp. 83–88)
D. Secular εὔχομαι introducing, concluding, and reporting speech (pp. 89–97)

      Classified list of attestations (p. 89):
          I: Direct speech: 1. Introducing speech; 2. Concluding speech
          II: Reported speech: 1. Absolute usage; 2. With aorist infinitive
                    a. Discussion of Section I attestations (pp. 89–97)
                    b. Discussion of Section II attestations (p. 97)
E. Conclusion (pp. 98–99) {68|69}

A. γένος εὔχομαι εἶναι and its transformations

I. H|| γένος εὔχομαι εἶναι #: Place or father

ξ 204 Κάστωρ Ὑλακίδης, τοῦ ἐγὼ γένος εὔχομαι εἶναι·
Ξ 113 πατρὸς δ’ ἐξ ἀγαθοῦ καὶ ἐγὼ γένος εὔχομαι εἶναι
ρ 373 αὐτὸν δ’ οὐ σάφα οἶδα, πόθεν γένος εὔχεται εἶναι
HApoll 470 εἰς Πύλον ἐκ Κρήτης, ἔνθεν γένος εὐχόμεθ’ εἶναι

II. γένος εὔχομαι (+/- ἔμμεναι): Place or father

ξ 199 ἐκ μὲν Κρητάων γένος εὔχομαι εὐρειάων
π 62 ἐκ μὲν Κρητάων γένος εὔχεται εὐρειάων
*φ 355 πατρὸς δ’ ἐξ ἀγαθοῦ γένος εὔχεται ἔμμεναι υἱός
*ω 269 εὔχετο δ’ ἐξ Ἰθάκης γένος ἔμμεναι, αὐτὰρ ἔφασκε

III. | εὔχομαι εἶναι #: Place or father

ο 425 ἐκ μὲν Σιδῶνος πολυχάλκου εὔχομαι εἶναι
Φ 187 αὐτὰρ ἐγὼ γενεὴν μεγάλου Διὸς εὔχομαι εἶναι
Υ 241 = Ζ 211 ταύτης τοι γενεῆς τε καὶ αἵματος εὔχομαι εἶναι
α 180 Μέντης Ἀγχιάλοιο δαΐφρονος εὔχομαι εἶναι # υἱός
HApoll 480 εἰμὶ δ’ ἐγὼ Διὸς υἱός, Ἀπόλλων δ’ εὔχομαι εἶναι
α 418 Μέντης Ἀγχιάλοιο δαΐφρονος εὔχεται εἶναι # υἱός
Ε 246 Πάνδαρος, υἱὸς δ’ αὖτε Λυκάονος εὔχεται εἶναι
α 406 ὁππόθεν οὗτος ἀνήρ· ποίης δ’ ἐξ’ εὔχεται εἶναι # γαίης
υ 192 ἡμέτερον πρὸς δῶμα; τέων δ’ ἐξ’ εὔχεται εἶναι
ι 519 τοῦ γὰρ ἐγὼ πάϊς εἰμί, πατὴρ δ’ ἐμὸς εὔχεται εἶναι
ι 529 εἰ ἐτεόν γε σός εἰμι, πατὴρ δ’ ἐμὸς εὔχεαι εἶναι
HHerm 378 πείθεο· καὶ γὰρ ἐμεῖο πατὴρ φίλος εὔχεαι εἶναι
*Ν 54 Ἕκτωρ, ὃς Διὸς εὔχετ’ ἐρισθενέος πάϊς εἶναι


Ε 247 Αἰνείας δ’ υἱὸς μὲν ἀμύμονος Ἀγχίσαο
         εὔχεται ἐκγεγάμεν, μήτηρ δέ οἵ ἐστ’ Ἀφροδίτη
Υ 208 αὐτὰρ ἐγὼν υἱὸς μεγαλήτορος Ἀγχίσαο
         εὔχομαι ἐκγεγάμεν, μήτηρ δέ μοί ἐστ’ Ἀφροδίτη

a. Rationale of classification

The preceding list contains all the Homeric attestations of secular εὔχομαι in the context of γένος specification. The εὔχομαι formulas in the three principal sections are apparently modifications of each other. For the relationship between

I.           Η|| γένος εὔχομαι εἶναι #


II.                     | εὔχομαι εἶναι #

compare the following:

(5 x)                     | ὠκὺς Ἀχιλλεύς # {69|70}
(9 x)           H|| πόδας ὠκέα Ἶρις #
(2 x)                     | ὠκέα Ἶρις #
(2 x)           H|| ὑπέλυσε δὲ γυῖα #
(8 x)                     | λῦσε δὲ γυῖα #
(1 x)           H|| ὑπέλυντο δὲ γυῖα #
(2 x)                     | λῦντο δὲ γυῖα #

Given the phraseological correspondences across these pairs, Nagy concludes that the switch in them is productive, i.e. it is a synchronically functioning mechanism for the epic poet. [1] However the relationship between

I.           H|| γένος εὔχομαι εἶναι #

and the expressions in Section II is not apparent from a synchronic perspective. Instead, a historical explanation is at hand in Nagy’s etymology of the dactylic hexameter. Let us examine the attestations in this section individually:

ξ 199 ἐκ μὲν Κρητάων γένος εὔχομαι εὐρειάων
π 62 ἐκ μὲν Κρητάων γένος εὔχεται εὐρειάων

The phrase γένος εὔχομαι, which occurs in only these two interdependent lines, is, I believe, an archaism. Such is immediately suggested by the grammatical archaisms which surround it. The phrase Κρητάων … εὐρειάων # is not only morphologically archaic in Homer, where the uncontracted α-stem genitive plural is metrically restricted, but also a syntactic archaism: these lines provide the only examples in existence of pluralized Crete. Certain place names of importance in Bronze Age Greece (Ἀθῆναι, Μυκῆναι, Ἀλαλκομεναί) occur in the plural, and this dead syntactic mechanism has been convincingly explained (also by G. Nagy, Lecture, 1969) as a function of political expansion in that period. The plural form of Ἀθήνη (attested once in the singular, η 80) [
2] means ‘Athens and its environs’, a designation applicable after the συνοικσμός of Attica: so also Κρηταί … εὐρεῖαι in the time of its political expansion. [3]

To return to || γένος εὔχομαι, its metrical shape is the same as that of the extremely archaic poetic phrase κλέος ἄφθιτον. In arguments to which I refer the reader, Nagy has shown that this phrase, which has an exact cognate in Vedic śrávasákṣitam,was fixed in Indo-European times in {70|71} the cadence of a lyric meter antecedent to the dactylic hexameter, as attested in Sappho 44.4 L–P (glyconics expanded by two dactyls): [4]

                                        … κλέος ἄφθιτον #

It adapted to the dactylic hexameter with the phraseological elaboration

I 413                                 … κλέος ἄφθιτον ἔσται #

Compare also:

Σ 2                                    … H|| ταχὺς ἄγγελος ἦλθε #
Sappho 44.3 L – P                       τάχυς ἄγγελος #

The latter lyric formula also occurs in Epic in the thrice-attested metrical position (Ω 292, Ω 310, o 526

Ω292, Ω310 || ταχὺν ἄγγελον, ὅς τε οἱ/σοι αὐτῷ
Ο526 || ταχὺς ἄγγελος, ἐν δὲ πόδεσσι

We can thus constitute a series with hypothetical members as follows:

Lyric:                                         | κλέος ἄφθιτον #
Epic:                               H|| κλέος ἄφθιτον ἔσται #
*Epic:           || κλέος ἄφθιτον
Lyric:                                         | ταχὺς ἄγγελος #
Epic:                               H|| ταχὺς ἄγγελος ἦλθε #
Epic:             || ταχὺν ἄγγελον
*Lyric:                                        | γένος εὔχομαι #
Epic:                               H|| γένος εὔχομαι εἶναι #
Epic:             || γένος εὔχομαι

The other line in Section II which attests || γένος εὔχομαι is

φ 335 πατρὸς δ’ ἐξ’ ἀγαθοῦ γένος εὔχεται ἔμμεναι υἱός

It is possible that the second part of this line preserves an archaic expansion of the phrase γένος εὔχομαι εἶναι #. For the relationship between them, consider the following series:

Ι 109                               H|| μεγαλήτορι θυμῷ #
O 674                     || μεγαλήτορι ἥνδανε θυμῷ #
υ 387                               H|| περικαλλέα δίφρον #
Γ 262                     || περικαλλέα βήσετο δίφρον #
passim                              H|| κορυθαίολος Ἕκτωρ #
Χ 471                     || κορυθαίολος ἠγάγεθ’ Ἕκτωρ #

Contrast this with:


While the phraseological correspondence here is not as simple as in the preceding, productive series, the metrical relationship between Η|| γένος εὔχομαι εἶναι # (pherecratic) and || γένος εὔχεται ἔμμεναι υἱός # (pherecratic {71|72} expanded by a dactyl) is the same. [
5] A clearer phraseological correspondence is attested twice in the enjambed formula (sic) | εὔχομαι εἶναι # υἱός (α 180, α 418). Perhaps this expression was once a lyric formula with the shape and position *υἱός εὔχομαι εἶναι # (pherecratic) which was ultimately replaced by the hexametric H|| – γένος εὔχομαι εἶναι # (also pherecratic). So there once would have been a pair:

                                          # *υἱὸς εὔχομαι εἶναι # (pher.)
φ 335                     || γένος εὔχεται ἔμμεναι υἱός # (pher.d)

As for the last attestation in Section II, it will be explained below in terms of doublet composition, where lexical and metrical divergence is the norm (see below, p. 76). Its language is, however, comparable to that of φ 335:

φ 335 πατρὸς δ’ ἐξ ἀγαθοῦ γένος εὔχεται ἔμμεναι υἱός
ω 269 εὔχετο δ’ ἐξ Ἰθάκης γένος ἔμμεναι, αὐτὰρ ἔφασκεν

and it is worth mentioning in this regard that the attestations in II are contextually interrelated. All four concern Odysseus’ genealogy, the first three when he is disguised as a beggar, this last when he tests Laertes and pretends he was Odysseus’ friend. Common archaisms are more easily preserved in common contexts.

The lines in all these sections, then, are synchronic or diachronic transformations of each other, but they are not such as to necessitate semantic distinctions in εὔχομαι between the sections. In their variety we can see the poet’s ways of adapting the idea γένος εὔχομαι εἶναι to different phraseological environments as well as the conservatism of his medium and memory. As another instance of this and as a sign of functional consistency in εὔχομαι, it is worthwhile noticing that there are strong grammatical constraints in force among all the lines (24). εὔχομαι occurs only once outside of the present (ω 269), once outside of the singular (HApoll 470), and just twice in the second person (ι 529, HHerm 378). [6] Functional fixity is {72|73} a fortiori suggested for the expressions in III, where the phraseological transformation H|| γένος εὔχομαι εἶναι # <=> | εὔχομαι εἶναι is a mechanism synchronically available to the poet. As a parallel, consider that there is no reason to believe that the word κορυθαίολος has a different meaning in

                              H|| κορυθαίολος Ἕκτωρ #

than it does in

                              || κορυθαίολος ἠγάγεθ’ Ἕκτωρ #

or that

                              | λῦσε δὲ γυῖα #

necessarily means something else than

                              H|| ὑπέλυσε δὲ γυῖα #

in spite of the lexical change ὑπέλυσε/λῦσε. Any discernible semantic difference between these expressions is strictly speaking a secondary phenomenon. [
7] How much less reason is there to suppose semantic differentiation between H|| γένος εὔχομαι εἶναι # and | εὔχομαι εἶναι # ?

Moreover, the wider contexts of all these lines are predictable and consistent, a fact which enhances the hypothesis of their functional consistency. All twenty-one lines in I, II, and III refer to, ask for or, most commonly, specify the place of origin or the male parent of a personage in the poem, divine or human. That this is an either/or phenomenon, either male parent or place, for each line is, again, not symptomatic of a functional split in εὔχομαι but of an artificial split in the category of γένος. In those contexts which allow for elaboration, the poet always gives both place of origin and person/s of origin. The classic example is Glaukos’ speech about his γενεή.

Ζ 145 Τυδεΐδη, μεγάθυμε, τίη γενεὴν ἐρέεινεις; …
          εἰ δ’ ἐθέλεις καὶ ταῦτα δαήμεναι, ὄφρ’ ἐῢ εἰδῇς
          ἡμετέρην γενεήν, πολλοὶ δέ μιν ἄνδρες ἴσασιν

where the genealogy begins as follows:

Ζ 152 ἔστι πόλις Ἐφύρη μυχῷ Ἄργεος ἱπποβότοιο,
          ἔνθα δὲ Σίσυφος ἔσκεν ὃ κέρδιστος γένετ’ ἀνδρῶν,
          Σίσυφος Αἰολίδης· ὁ δ’ ἄρα Γλαῦκον τέκεθ’ υἱόν,
          αὐτὰρ Γλαῦκος τίκτεν ἀμύμονα Βελλεροφόντην … {73|74}

Ζ 211 ταύτης τοι γενεῆς τε καὶ αἵματος εὔχομαι εἶναι

Or compare Asteropaios’ genealogy:

Φ 153 Πηλεΐδη μεγάθυμε, τίη γενεὴν ἐρεείνεις;
          εἴμ’ ἐκ Παιονίης ἐριβώλου, τήλυθ’ ἐούσης,
          κ. τ. λ.

or that of Aeneas

Υ 213 εἰ δ’ ἐθέλεις καὶ ταῦτα δαήμεναι, ὄφρ’ ἐῢ εἰδῇς
          ἡμετέρην γενεήν, πολλοὶ δέ μιν ἄνδρες ἴσασι·
          Δάρδανον αὖ πρῶτον τέκετο νεφεληγερέτα Ζεύς,
          κτίσσε δὲ Δαρδανίην, ἐπεὶ οὔ πω Ἴλιας ἱρὴ
          ἐν πεδίῳ πεπόλιστο, πόλις μερόπων ἀνθρώπων,
          ἀλλ’ ἔθ’ ὑπωρείας ᾤκεον πολυπίδακος Ἴδης.
          Δάρδανος αὖ τέκεθ’ υἱὸν Ἐριχθόνιον βασιλῆα,
          κ. τ. λ.

In these formal and mannered speeches, a hero has space to give his γενεή–not only his parents, their parents, etc. and their activities, but also the locale of his generation. [
8] However, compendious specification of a person’s γένος, such as the single lines we are discussing demand, has space only for the name of an immediate male ancestor or a toponym with epithet. We can adduce a typological parallel. Both place of origin and father’s name are regular onomastic categories (e.g. Robertson, Michaelson, Johnson or Boston Blackie, Ludwig van Beethoven, Nathan Detroit), but specification of one or the other in names is a permissible norm. This analogy to onomastic categories is apt, since the heroes or gods in these twenty-one lines are actually naming themselves or being named in terms of epic conventions. [9]  {74|75}

Finally, there are two attestations of εὔχομαι in the context of γένος specification which are, phraseologically speaking, widely divergent. These are the doublets about Aeneas: [10]

Ε 247 Αἰνείας δ’ υἱὸς μὲν ἀμύμονος Ἀγχίσαο
          εὔχεται ἐκγεγάμεν, μήτηρ δέ οἵ ἐστ’ Ἀφροδίτη.
Υ 208 αὐτὰρ ἐγὼν υἱὸς μεγαλήτορος Ἀγχίσαο
          εὔχομαι ἐκγεγάμεν, μήτηρ δέ μοί ἐστ’ Ἀφροδίτη

The reasons for doublet composition of Aeneas’ γένος are parallel to those for doublet composition of reported prayers (see above, p. 65). Exceptionally, the most important figure in his γένος is his mother, who is a goddess. Given the convention of patronymics for all or metronymics for those descended from gods/goddesses, the tradition provides both Aeneas’ father’s and his mother’s name. It adjusts formulaic language more radically than elsewhere (# εὔχομαι ἐκγεγάμεν ~ γένος εὔχομαι εἶναι #) and provides all the information in a gross formula of two full lines’ extent. [
11] So compendious a statement of an idea greater than formula length (father’s +/- mother’s name) results in unique lexical features which the poet can only reproduce once if at all. The use of this compositional tool supports our {75|76} previous contention that the place/father’s name split is motivated by technical expedience and not a functional difference in εὔχομαι. Specification of both place and father’s name for each hero would result in a large number of complex, ad hoc formulas. The mechanics of composition also suggest a more precise explanation for the less complex transformation of H|| # γένος εὔχομαι εἶναι # in III and for the preservation of archaisms in II. A shorter formula, such as | εὔχομαι εἶναι # is more useful to the poet than a longer one (e.g. the doublet, which is a large unit of specialized language difficult to regenerate), and, in general, the mechanism for γένος specification is not inflexible since place names and father’s names have a wide variety of metrical shapes. (Contrast, for example, the relative consistency of metrical shapes for the three gods prayed to in the dative, above, p. 47). Occasionally, as in II, the result is metrical displacement accompanied by contraction or expansion of a formula H|| γένος εὔχομαι εἶναι #. We can see that Homeric reality is often but not always susceptible to symmetrical and rigid formula systems. Conversely, the traditional language is resilient and equipped to express poetically the multiplicity of that reality. It is not banal to insist, by way of conclusion, that the form and context of all the εὔχομαι expressions in I–IV imply that the word’s function in them is unitary: the poet can achieve his end by making or preserving minimal changes within his inherited repertoire.

b. The meaning of εὔχομαι:

However, the poet has one more option, and it is an option which has capital importance for a study of the meaning of εὔχομαι in these attestations: he can use the word φημί as a substitute or synonym for εὔχομαι. Consider, for example, the following lines:

ω 269 εὔχετο δ’ ἐξ’ Ἰθάκης γένος ἔμμεναι, αὐτὰρ ἔφασκεν
          Λαέρτην Ἀρκεισιάδην πατέρ’ ἔμμεναι αὐτῷ.

By dividing the γένος statement into two parts, place of origin and father’s name, and introducing one with εὔχετο and the other with the iterative imperfect ἔφασκε, the poet produces a potential doublet in his desire to give complete but compendious information. [
12] By itself, this attestation implies much about the meaning of εὔχομαι. φημί, we know, is the general or functionally unmarked word for speech of any kind. It follows from its functional and formal parallelism to εὔχομαι in these lines that the latter is simply a marked word for speech. ‘Say’ can substitute for it, or, in certain contexts, contrast with it:

Υ 206 φασὶ σὲ μὲν Πηλῆος ἀμύμονος ἔκγονον εἶναι
          μητρὸς δ’ ἐκ Θέτιδος καλλιπλοκάμου ἁλοσύδνης, {76|77}
          αὐτὰρ ἐγὼν υἱὸς μεγαλήτορος Ἀγχίσαο
          εὔχομαι ἐκγεγάμεν, μήτηρ δέ μοί ἐστ’ Ἀφροδίτη

Here Aeneas is contrasting his own origin from an Olympian mother with Achilles’ from a sea-nymph. [
13] For Achilles he uses the unmarked φημί in the unmarked third person plural, while for himself, Aeneas uses εὔχομαι, first person singular. [14] Again, when Achilles wishes to intimidate a hero of lesser stature than himself, he employs the same lexical contrast:

Φ 186 φῆσθα σὺ μὲν ποταμοῦ γένος ἔμμεναι εὐρὺ ῥέοντος,
          αὐτὰρ ἐγὼ γενεὴν μεγάλου Διὸς εὔχομαι εἶναι.

By referring to his γενεή (‘long-range lineage or complete ancestry’) as opposed to his γένος (‘immediate ancestry’), [
15] he increases the contrast between himself and Asteropaios. This in turn enables him to use the pointed φημί – εὔχομαι contrast. [16] Compare the disequilibrium in these passages to the parallelism in the following one:

Ε 243 Τυδεΐδη Διόμηδες, ἐμῷ κεχαρισμένε θυμῷ,
          ἄνδρ’ ὁρόω κρατερὼ ἐπὶ σοὶ μεμαῶτε μάχεσθαι,
          ἶν’ ἀπέλεθρον ἔχοντας· ὁ μὲν τόξων ἐῢ εἰδώς,
          Πάνδαρος υἱὸς δ’ αὖτε Λυκάονος εὔχεται εἶναι·
          Αἰνείας δ’ υἱὸς μὲν ἀμύμονος Ἀγχίσαο
          εὔχεται ἐκγεγάμεν, μήτηρ δέ οἵ ἐστ’ Ἀφροδίτη.

Here two famous Trojan warriors are approaching the hard-pressed Greeks, and Sthenelus, who is trying to convince Diomedes they should flee, uses εὔχεται for both of them. He is not contrasting Pandaros and Aeneas, but rather their combined prestige with his own and Diomedes’ power to resist. Conversely, when Telemachus is self-destructively modest about his γένος, he consistently uses the verb φημί: {77|78}

Note that φημί can be active and have a different subject than its infinitive. In fact, its subject is often unmarked: φασι = on dit. [
18] εὔχομαι is by polar contrast a literally egocentric word, since its subject and that of its infinitive are always specific and always the same. These are all aspects of its usage which strongly support a definition of εὔχομαι as a functionally marked word for ‘say’. To go a step further and define it as ‘boast’, which all the handbooks do, is to obscure its living semantic link with φημί. The dimensions of its markedness are indeed superiority and contentiousness, but they are not basic to its meaning. In addition, another dimension to its marked sphere of reference is accuracy, for no one εὔχεται a false or speculative γένος or γενεή. This is another reason for Telemachus’ reluctance to say εὔχομαι in the passage just quoted (α 215ff.), and it gives a greater impact to the only line where a hero is not given his expected parentage:

Ν 54 Ἕκτωρ, ὃς Διὸς εὔχετ’ ἐρισθενέος πάϊς εἶναι

This, the ultimate genealogy, is being put forward not as pretentious or boastful but true. [
19] In terms of current lexicographical conventions, then, I suggest that the definition of εὔχομαι in this usage should read ‘say (proudly, accurately, contentiously)’. [20]  {78|79}

B. εὔχομαι εἶναι + comparative/superlative

I. εὔχομαι εἶναι + comparative

a) Age

Ι 161 ἠδ’ ὅσσον γενεῇ προγενέστερος εὔχομαι εἶναι
γ 362 οἶος γὰρ μετὰ τοῖσι γεραίτερος εὔχομαι εἶναι
Ι 60 ἀλλ’ ἄγ’ ἐγών, ὃς σεῖο γεραίτερος εὔχομαι εἶναι
b) Physical Prowess, Physical Beauty, Prestige
Δ 405 ἡμεῖς τοι πατέρων μέγ’ ἀμείνονες εὐχόμεθ’ εἶναι
Ε 173 οὐδέ τις ἐν Λυκίῃ σέο γ’ εὔχεται εἶναι ἀμείνων
ε 211 οὐ μέν θην κείνης γε χερείων εὔχομαι εἶναι
*Φ 410 νηπύτι’, οὐδέ νύ πώ περ ἐπεφράσω ὅσσον ἀρείων
            εὔχομ’ ἐγὼν ἔμεναι, ὅτι μοι μένος ἰσοφαρίζεις

II. εὔχομαι εἶναι + ἄριστος: Prestige, Physical Prowess

Α 91 ὃς νῦν πολλὸν ἄριστος Ἀχαιῶν εὔχεται εἶναι
Β 82 νῦν δ’ ἴδεν ὃς μέγ’ ἄριστος Ἀχαιῶν εὔχεται εἶναι
Ο 296 αὐτοὶ δ’ ὅσσοι ἄριστοι ἐνὶ στρατῷ εὐχόμεθ’ εἶναι
Ψ 669 πυγμῇ νικήσαντ’ ἐπεὶ εὔχομαι εἶναι ἄριστος
Υ 102 νικήσει’, οὐδ’ εἰ παγχάλκεος εὔχεται εἶναι

a. Rationale of classification

These lines contain the same expression as occurs in Section III. of the γένος εὔχομαι εἶναι attestations, but its context differs, and, concurrently, it can undergo a positional movement not exhibited by those lines, namely

                    | εὔχομαι εἶναι # ~ εὔχεται | εἶναι

In Hainsworth’s lists of mobile formulae with this shape, I count twenty-eight which move between these positions, including such expressions as πατρίδα γαῖαν (62 times line-final, 15 times crossing the diaeresis), οὐρανὸν εὐρύν (25, 7), νόστιμον ἦμαρ (4, 7), etc. [
21] In view of the commonness of this switch, the first question to ask is not why does εὔχομαι εἶναι move in these contexts, but rather, why is its position fixed in the γένος context? The answer must be that metrical position and semantic context are in this case interdependent. εὔχομαι εἶναι # specifying a person’s γένος has its positional rigidity conditioned by the positional rigidity of the expression {79|80} γένος εὔχομαι εἶναι # , which is frozen at the cadence of the line and which is a synchronic variant of it (see above, p. 70). [22] When the context of εὕχομαι εἶναι changes from γένος specification, the formula’s synchronic bonds to γένος εὔχομαι εἶναι # are loosened and it becomes an independent, mobile expression. [23] This is a minimalistic explanation of a formal phenomenon on the relatively formal grounds of contextual change. To go further and suppose that the function of εὔχομαι undergoes a change is not warranted by the mobility of this formula unless the contextual change is incompatibly great. That is doubtful since the formula itself (εὔχομαι εἶναι) remains intact, but it is a possibility we will consider below. In one instance, it is true, the formula does undergo lexical change, but this is only because it is enjambed to its comparative,

Φ 410                                        ὃσσον ἀρείων #
                 εὔχομ’ ἐγὼν ἔμεναι …

an occurrence which results in x-sector improvisational phraseology. [
24] Moreover, the lexical and metrical variety of the comparatives and the superlative with which εὔχομαι εἶναι is associated (|| προγενέστερος, || γεραίτερος [2 x], ἀρείων #, ἀμείνων #, ἀμείνονες|, χερείων|, ἄριστος/οι || [3 x], ἄριστος #) implies that εὔχομαι εἶναι is contextually but not qua formula fused with any except perhaps the superlative ἄριστος and || γεραίτερος. But rather than atomize these attestations—an operation which could also be performed upon the attestations in Section III of the γένος εὔχομαι εἶναι formulas—it seems better to consider the whole group as a system whose character resembles the infinitive + ἀνώγει/ε/α group cited above (see the list, Ch. II., p. 52) where lexical consistency is occasionally the result of a phraseological pattern. In other words, we have here a ‘generative’ formula, to use Nagler’s terminology, or a formula system as Parry defined it. [25] The best way to label it is “comparative/superlative + εὔχομαι εἶναι.” This kind of formulation is ultimately another sign of the epicene status of εὔχομαι εἶναι in this context vis-à-vis the γένος context, a sign which we can correlate with its mobility. Both phenomena imply the quasi-independence of this εὔχομαι εἶναι from γένος εὔχομαι εἶναι # and Section III’s εὔχομαι εἶναι #, but they do not yet warrant a functional divorce between the two usages. {80|81}

b. The meaning of εὔχομαι

The preceding formal analysis is weak unless the contextual unity of these εὔχομαι εἶναι expressions is clear and their relationship to the unified γένος εὔχομαι εἶναι contexts is textually ascertained. The only possible obstacle to the first proposition is the apparent gap between statements of superiority in age and statements of superiority in prestige, physical prowess, and physical beauty, the unity of these three being a Homeric truism. But this is no gap at all. In Homer, age is an index of prestige. Consider the context in which, for example, Agamemnon utters the first attestation of εὔχομαι εἶναι on the list. He is giving the embassy to Achilles its charge, and in spite of the loss of prestige for him that such an embassy involves, he wants Achilles to accept the ἄποινα as a token of his, Agamemnon’s, prestige, under threat of death: [26]

Ι 158 δμηθήτω· — Ἀΐδης τοι ἀμείλιχος ἠδ’ ἀδάμαστος,
          τοὔνεκα καί τε βροτοῖσι θεῶν ἔχθιστος ἁπάντων —
          καί μοι ὑποστήτω, ὅσσον βασιλεύτερός εἰμι
          ἠδ’ ὅσσον γενεῇ προγενέστερος εὔχομαι εἶναι.

Agamemnon is here asserting his superiority in prestige to Achilles in the same language even with which Achilles compares his lineage to Asteropaios’ and asserts his superiority:

Φ 186 φῆσθα σὺ μὲν ποταμοῦ γένος ἔμμεναι εὐρὺ ῥέοντος
          αὐτὰρ ἐγὼ γενεὴν μεγάλου Διὸς εὔχομαι εἶναι.

Moreover, this is the same Agamemnon of whom Achilles uses the word ἄριστος in Book A:

Α 90                               …οὐδ’ ἤν Ἀγαμέμνονα εἴπῃς #
          ὃς νῦν πολλὸν ἄριστος Ἀχαιῶν εὔχεται εἶναι.

So also Nestor, speaking of Agamemnon’s vision of the false dream and commending its message to the βουλή of chieftains in Book B:

Β 82 νῦν δ’ ἴδεν ὃς μέγ’ ἄριστος Ἀχαιῶν εὔχεται εἶναι.

This set of passages, with Ι 161 as a point of departure, attests to the close association in Homer of age, prestige, and also birth. It therefore supports the functional identity of εὔχομαι εἶναι in both γένος contexts and the contexts of these comparatives and superlatives, which are, in fact, barely distinguishable. For more contextual support, compare the conclusion of Glaukos’ famous speech about his γενεή:

Ζ 206 Ἱππόλοχος δέ μ’ ἔτικτε, καὶ ἐκ τοῦ φημι γενέσθαι·
          πέμπε δέ μ’ ἑς Τροίην, καί μοι μάλα πόλλ’ ἐπέτελλεν
          αἰὲν ἀριστεύειν καὶ ὑπείροχον ἔμμεναι ἄλλων {81|82}
          μηδὲ γένος πατέρων αἰσχυνέμεν, οἳ μέγ’ ἄριστοι
          ἔν τ’ Ἐφύρῃ ἐγένοντο καὶ ἐν Λυκίῃ εὐρείῃ.
          ταύτης τοι γενεῆς τε καὶ αἵματος εὔχομαι εἶναι.

ἄριστος in this context means ‘noblest’, ‘best as a function of prestigious γενεή’. Compare to it Aeneas’ exhortation to Pandaros :

Ε 171 Πάνδαρε, ποῦ τοι τόξον ἰδὲ πτερόεντες ὀϊστοὶ
          καὶ κλέος, ᾧ οὔ τίς τοι ἐρίζεται ἐνθάδε γ’ ἀνήρ
          οὐδέ τις ἑν Λυκίῃ σέο γ’ εὔχεται εἶναι ἀμείνων;

where ἀμείνων refers to Pandaros’ physical skill in archery, which is, in turn, the substance of his κλέος, the word for glory and prestige which transcends mortality. The phrase εὔχεται εἶναι is not lightly used in this context. Pandaros has just fatally wounded Diomedes with an arrow, but miraculously Athena healed him and negated Pandaros’ skill. In disgust, he promises to burn his bow and arrows, then takes up a spear against Diomedes, but he misses and is killed by him. Thus the myth equates Pandaros’ failure and loss of faith in archery, brought about only by the miraculous intervention of a goddess, with his death. It is not difficult to conclude that Pandaros’ identity resides in his bow and arrows, and that if no one in Lycia εὔχεται to be better with them than he, there is good reason still. To sum up, the context of Ε 173 unites physical skill (archery), prestige (κλέος), and identity, and its language is reminiscent of the γενεή language in Glaukos’ speech (see above, Z 206–11).

C. | εὔχομαι εἶναι # +social relationships

I. ξεῖνος, ἱκέτης

ω 114 εἰπέ μοι εἰρομένῳ· ξεῖνος δέ τοι εὔχομαι εἶναι
ε 450 ἀλλ’ ἐλέαιρε ἄναξ· ἱκέτης δέ τοι εὔχομαι εἶναι
π 67 ἔρξον ὅπως ἐθέλεις· ἱκέτης δέ τοι εὔχεται εἶναι
Ζ 231 γνῶσιν ὅτι ξεῖνοι πατρώϊοι εὐχόμεθ’ εἶναι
α 187 ξεῖνοι δ’ ἀλλήλων πατρώϊοι εὐχόμεθ’ εἶναι
ο 196 μῦθον ἐμόν; ξεῖνοι δὲ διαμπερὲς εὐχομεθ’ εἶναι

II. Miscellaneous

Θ 190 ἢ ἐμοί, ὅς πέρ οἱ θαλερὸς πόσις εὔχομαι εἶναι
χ 321 εἰ μὲν δὴ μετὰ τοῖσι θυοσκόος εὔχεαι εἶναι
ι 263 λαοὶ δ’ Ἀτρεΐδεω Ἀγαμέμνονος εὐχομεθ’ εἶναι

Formally speaking, the metrical and lexical fixation of εὔχομαι εἶναι # in all these attestations calls into question their treatment as a set independent of the εὔχομαι εἶναι + # place/father’s name attestations cited {83|84} above (p. 69). Separate treatment seems least justifiable for the last three on the preceding list. In them, the metrical expediency of deleting γένος from the expression H|| γένος εὔχομαι εἶναι # has given a measure of contextual flexibility as well to the poet. | εὔχομαι εἶναι # now accompanies not specification of γένος but specification of a close social relationship to some important personage which is as distinctive a mark of identification and privilege as father’s name or fatherland. For Hector’s identification of himself as Andromache’s potent spouse

Θ 190 ἢ ἐμοί, ὃς πέρ οἱ θαλερὸς πόσις εὔχομαι εἶναι

compare ἐρίγδουπος πόσις Ἣρης # (passim, = Zeus) [
28] and also the lines

λ 236 ἣ φάτο Σαλμωνῆος ἀμύνονος ἔκγονος εἶναι,
         φῆ δὲ Κρηθῆος γυνὴ ἔμμεναι Αἰολίδαο

where forms of φημί substitute for εὔχομαι, and the specification of spouse accompanies γένος specification. In the second example, ι 263, Odysseus is evasively answering the Cyclops’ questions about his identity (ι 252–255). Full citation of the context shows that the function of εὔχομαι is again ‘say (proudly, accurately, contentiously)’:

ι 263 λαοὶ δ’ Ἀτρεΐδεω Ἀγαμέμνονος εὐχόμεθ’ εἶναι
         τοῦ δὴ νῦν γε μέγιστον ὑπουράνιον κλέος ἐστί
         τόσσην γὰρ διέπερσε πόλιν καὶ ἀπώλεσε λαοὺς

The reader should not be misled by the social status of λαός (‘army’) into interpreting this attestation as a demotion of εὔχομαι to simply ‘say’. As in the γένος εὔχομαι εἶναι # usage, the source of pride and prestige is not primarily the relationship specified (son, inhabitant, army of …) but the person or place to which one is related (father, place, leader). Thus the stress in this passage is on Agamemnon’s κλέος, not the army’s.

The usage of εὔχομαι εἶναι # in these three attestations, then, precisely parallels its usage in the εὔχομαι εἶναι # + place of birth/father’s name attestations. All are identity statements in which the meaning ‘say (proudly, accurately, contentiously)’ for εὔχομαι is to the point. At first blush, however, it is not immediately apparent that the other attestations in this group (of the type ξεῖνος/ ἱκέτης + εὔχομαι εἶναι #) are also identity statements. Nevertheless an examination of the wider contexts in which they occur shows this to be true. For example, when Glaukos concludes his γενεή-speech with the line

Z 211 ταύτης τοι γενεῆς τε καὶ αἵματος εὔχομαι εἶναι

Diomedes smiles, plants his spear in the ground, and sweetly tells Glaukos that he is his πατρώϊος ξεῖνος. One of Diomedes’ ancestors, it turns out, gave hospitality to the most famous of Glaukos’ ancestors, Bellerophon. This relationship is a source of public pride to Diomedes, as is apparent from the concluding lines of his speech:

Ζ 230 τεύχεα δ’ ἀλλήλοις ἐπαμείψομεν, ὄφρα καὶ οἳδε
γνῶσιν ὅτι ξεῖνοι πατρώϊοι εὐχόμεθ’ εἶναι.
(οἳδε = Achaeans and Trojans)

The formal (εὔχομαι εἶναι # vs. εὐχόμεθ’ εἶναι # ) and positional (speech conclusion) parallelism of Ζ 211 and Ζ 231 elucidates the εὔχομαι expression {85|86} in the latter. By ascertaining and publicizing this attachment to Glaukos, Diomedes enhances his prestige. The same applies in principle to Glaukos. And as the word πατρώϊοι – used here with a formula elsewhere associated with fathers and fatherlands – implies, the people with whom a hero has relations of hospitality form an integral part of his γενεή. In a word, ties of hospitality in this society are closely akin to ties of blood, and the function of εὔχομαι εἶναι # in both of these contexts is identical.

It’s not necessary to demonstrate this in detail for the remaining attestations. This same phrase, ξεῖνοι … πατρώϊοι εὐχόμεθ’ εἶναι occurs once again in the long answer Athena-Mentes gives to Telemachus’ long list of questions about her identity (α 170 – 177). First she gives her name and her father’s name in the line discussed above (n. 9):

α 180 Μέντης Ἀγχιάλοιο δαΐφρονος εὔχομαι εἶναι

and a few lines later she says that she and Odysseus are each other’s guest and host:

α 187 ξεῖνοι δ’ ἀλλήλων πατρώϊοι εὐχόμεθ’ εἶναι

Here again, γένος-specification parallels ξεῖνος-specification. Elsewhere the expression ξεῖνος δέ τοι εὔχομαι εἶναι is used twice (ο 196, ω 114) when a guest wants to justify asking his host a special favor. This usage is simply an extension of the previous one. Its purpose is not to identify the guest, but to reassert his identity in a way which is politely flattering to the host: “I am proud to say I am your guest-host … tell me such and such …” or “do me this favor …”

This is precisely parallel to the usage of the two remaining attestations of εὔχομαι εἶναι#:

ε 450 ἀλλ’ ἐλέαιρε ἄναξ· ἱκέτης δέ τοι εὔχομαι εἶναι
π 67 ἔρξον ὅπως ἐθέλεις· ἱκέτης δέ τοι εὔχομαι εἶναι

except the relationship specified is ἱκέτης, not ξεῖνος. But these two words are bound together in Epic language elsewhere, for example

θ 546 ἀντὶ κασιγνήτου ξεῖνός θ’ ἱκέτης τε τέτυκται
         ἀνέρι, …

where Alkinoos asserts the proximity of blood ties to those of guest/host and ἱκέτης, or

ι 266 … ἡμεῖς δ’ αὖτε κιχανόμενοι τὰ σὰ γοῦνα
         ἱκόμεθ’, εἴ τι πόροις ξεινήϊον ἠὲ καὶ ἄλλως
         δοίης δωτίνην, ἥ τε ξείνων θέμις ἐστίν.
         ἀλλ’ αἰδεῖο, φέριστε, θεούς· ἱκέται δέ τοί εἰμεν.
         Ζεὺς δ’ ἐπιτιμήτωρ ἱκετάων τε ξείνων τε,
         ξείνιος, ὃς ξείνοισιν ἅμ’ αἰδοίοισιν ὀπηδεῖ.


η 162 ἀλλ’ ἄγε δὴ ξεῖνον μὲν ἐπὶ θρόνου ἀργηροήλου
         εἷσον ἀναστήσας, σὺ δὲ κηρύκεσσι κέλευσον {86|87}
         οἶνον ἐπικρῆσαι, ἵνα καὶ Διὶ τερπικεραύνῳ
         σπείσομεν, ὅς θ’ ἱκέτῃσιν ἃμ’ αἰδοίοισιν ὀπηδεῖ.
         δόρπον δε ξείνῳ ταμίη δότω ἔνδον ἐόντων.

These passages show that there is a functional overlap between ἱκέτης and ξεῖνος which is not apparent from their accepted meanings ‘suppliant’ and ‘guest-host’. The meaning of ἱκέτης needs clarification. Benveniste has recently defended its old etymology as nom d’agent of ἱκάνω, ἵκομαι (root ἱκ-) on the basis of figurae etymologicae in Homer (e.g. ε 445–450) and the pregnant meaning of the verb (‘atteindre, toucher’) in its less banal usages. [
31] He considers the frequent contextual liaison of ἱκέσθαι and γούνατα (‘parvenir aux genoux’) as the origin of the meaning ‘suppliant’ for ἱκέτης, as in

ε 449 σόν τε ῥόον σά τε γούναθ’ ἱκάνω πολλὰ μογήσας
         ἀλλ’ ἐλέαιρε ἄναξ· ἱκέτης δέ τοι εὔχομαι εἶναι.

This is a convincing explanation for the meaning ‘suppliant’. However, one may wonder on the basis of this evidence and the collocation of ξεῖνος and ἱκέτης whether the Homeric word ἱκέτης has actually developed the sense ‘suppliant’. I suggest that it still means ‘he who comes, arrives’. By the plain fact of his arrival in someone else’s house, the ἱκέτης can place an obligation on his host to give him gifts and offer hospitality—an obligation which, according to a passage just cited (ι 266) cannot be refused without incurring the wrath of Zeus ξείνιος. [
32] This seems a strict code of behavior, but its strictness is by no means unparalleled. The “obligation to repay gifts received” which is implicit in the Homeric institution of ξεινία has, in other archaic societies, two corollaries:

Likewise in Homeric society a person is constrained to give hospitality to anyone who arrives (ἱκάνω, ἵκω) at his house: {87|88}

π 78 ἀλλ’ ἦ τοι τὸν ξεῖνον, ἐπεὶ τεὸν ἵκετο δῶμα.
        ἕσσω μιν χλαῖνάν τε χιτῶνά τε, εἵματα καλά …
φ 312 = υ 294 … οὐ μὲν καλὸν ἀτέμβειν οὐδὲ δίκαιον
          ξείνους Τηλεμάχου, ὅς κεν τάδε δώμαθ’ ἵκηται.
γ 352 οὔ θην δὴ τοῦδ’ ἀνδρὸς Ὀδυσσῆος φίλος υἱὸς
         νηὸς ἐπ’ ἰκριόφιν καταλέξεται, ὄφρ’ ἂν ἐγώ γε
         ζώω, ἔπειτα δὲ παῖδες ἐνὶ μεγάροισι λίπωνται
         ξείνους ξεινίζειν, ὅς τίς κ’ ἐμὰ δώμαθ’ ἵκηται.

Notice that the collocation of ἱκάνω/ἵκω and ξεῖνος in this passage and others (e.g. η 24, π 57, θ 28) parallels the earlier collocations of ἱκέτης and ξεῖνος and reinforces our hypothesis that the semantic relationship between ἱκάνω/ἵκω and ἱκέτης is not only etymological but still felt.

To return to the εὔχομαι – ἱκέτης attestations, we can now suggest that the only difference between, e.g.

ω 114 εἰπέ μοι εἰρομένῳ· ξεῖνος δέ τοι εὔχομαι εἶναι


ε 450 ἀλλ’ ἐλέαιρε ἄναξ· ἱκέτης δέ τοι εὔχομαι εἶναι

is that the speaker of the second line has yet to receive hospitality, while that of the first may or may not have (in fact, he has, but the word ξεῖνος is ambiguous, whence the definitions ‘guest’, ‘host’, ‘stranger’). An ἱκέτης is simply ‘one who comes’, which in the formal and deceptive language of reciprocity means a ξεῖνος in need of his first favor. This usage is the point of departure for the adaption of ἱκέσθαι to its liaison with γούνατα, and it clarifies the functional overlap between ξεῖνος (unmarked) and ἱκέτης (marked). Accordingly, the phrase ἱκέτης δέ τοι εὔχομαι εἶναι # does not signal a demotion of εὔχομαι’s function to, e.g. ‘I say (humbly) that I am your suppliant’. It is simply a special case of ξεῖνος δέ τοι εὔχομαι εἶναι # to which the meaning ‘say (proudly, accurately, contentiously) that I am your visitor’. [
34] For the meaning ‘say’ in both usages, compare

ρ 522 φησὶ δ’ Ὀδυσσῆος ξεῖνος πατρώϊος εἶναι
τ 191 ξεῖνον γάρ οἱ ἔφασκε φίλον τ’ ἔμεν αἰδοῖόν τε {88|89}

D. Secular εὔχομαι introducing, concluding, and reporting speech:

I. Direct speech:

1. Introducing speech

Ν 619 τεύχεά τ’ ἐξενάριξε καὶ εὐχόμενος ἔπος ηὔδα
Ρ 537 τεύχεά τ’ ἐξενάριξε καὶ εὐχόμενος ἔπος ηὔδα
Φ 183 τεύχεά τ’ ἐξενάριξε καὶ εὐχόμενος ἔπος ηὔδα
Ξ 500 πέφραδέ τε Τρώεσσι καὶ εὐχόμενος ἔπος ηὔδα
Υ 424 ὡς εἶδ’, ὣς ἀνεπᾶλτο, καὶ εὐχόμενος ἔπος ηὔδα
Λ 379 ἐκ λόχου ἀμπήδησε καὶ εὐχόμενος ἔπος ηὔδα
*Τ 100 ἤτοι ὅ γ’ εὐχόμενος μετέφη πάντεσσι θεοῖσι
*ξ 463 εὐξάμενός τι ἔπος ἐρέω· οἶνος γὰρ ἀνώγει

2. Concluding speech

Ν 417 ὣς ἔφατ’ Ἀργείοισι δ’ ἄχος γένετ’ εὐξαμένοιο
Ξ 458 ὣς ἔφατ’ Ἀργείοισι δ’ ἄχος γένετ’ εὐξαμένοιο
Ξ 486 ὣς ἔφατ’ Ἀργείοισι δ’ ἄχος γένετ’ εὐξαμένοιο

II. Reported speech:

1. Absolute usage

Ξ 366 ἀλλ’ ὁ μὲν οὕτω φησὶ καὶ εὔχεται, οὕνεκ’ Ἀχιλλεύς
Α 397 εὐχομένης, ὅτ’ ἔφησθα κελαινεφέϊ Κρονίωνι
         οἴη ἐν ἀθανάτοισιν ἀεικέα λοιγὸν ἀμῦναι
Β 597 στεῦτο γὰρ εὐχόμενος νικησέμεν εἴ περ ἂν αὐταὶ
Ν 447 τρεῖς ἑνὸς ἀντὶ πεφάσθαι· ἐπεὶ σύ περ εὔχεαι οὕτω
Λ 388 νῦν δέ μ’ ἐπιγράψας ταρσὸν ποδὸς εὔχεαι αὔτως
Π 844 ἤδη νῦν Ἕκτορ μεγάλ’ εὔχεο · σοὶ γὰρ ἔδωκε

2. With aorist infinitive

Φ 501 εὔχεσθαι ἐμὲ νικῆσαι κρατερῆφι βίηφιν
Ξ 484 …τῶ καί τίς τ’ εὔχεται ἀνήρ
         γνωτὸν ἐνὶ μεγάροισιν ἀρῆς ἀλκτῆρα λιπέσθαι
*λ 261 ἣ δὴ καὶ Διὸς εὔχετ’ ἐν ἀγκοίνῃσιν ἰαῦσαι

a. Section I attestations

Following the method used above, we consider first the formulaic usages and then the ‘non-formulaic’ ones so as to employ the latter as a check on semantic hypotheses for the former and vice versa. The attestations comprise two formulas, one used (6 times) to introduce speech,

(X)                                         || καὶ εὐχόμενος ἔπος ηὔδα #

the other (3 times) to conclude it :

(Y)      ὣς ἔφατ’ Ἀργείοισι δ’ ἄχος γένετ’ εὐξαμένοιο

Both expressions are heavily restricted. Neither occurs outside of battle-books, and they are never used for the same speech. In fact, (X) introduces speech by a Greek, while (Y), as Ἀργείοισι δ’ ἄχος necessitates, concludes {89|90} speech by a Trojan. (There is one exception to this distributional rule that we will consider shortly.) There is therefore a functional polarization between (X) and (Y) which we can schematize as follows:

(X) – introduce speech – Greek vs. (Y) – conclude speech – Trojan.

The implication of this polarization is that the semantic function of εὔχομαι is identical in both (X) and (Y). It seems obvious that the difference in usage and formula shape between (X) and (Y) serves not to distinguish different kinds of speech but rather to clarify for the poet and his audience the allegiance of the diverse heroes who occur in battle-books and speak the speeches introduced or concluded by (X) and (Y). Notice that εὔχομαι is the only word which occurs in both (X) and (Y).

The restrictions go further. In every case except one – and the exception to this restriction (Λ 379) is the same as the exception to the preceding one – the speeches framed by (X) or (Y) are spoken by a warrior immediately after killing his enemy, Greek or Trojan in single combat. A schematization of the polarization now looks like this:

(X) – introduce speech – Greek speaker – Dead Trojan


(Y) – conclude speech – Trojan speaker – Dead Greek

It is possible that the purpose of this opposition is less mechanical than we at first supposed, since it constitutes a powerful metaphor for the deadly competition that takes place in battle-books, with every occurrence of an (X) calling up the expectation of a (Y) and vice versa. (For the effect of their alternation in a short passage, see Ξ 440-507). Also, the diversity of heroes is likely to be more confusing to our senses, which are unattuned to catalogue poetry, than to the senses of the traditional poet and his well-versed audience. It is in fact safer to assume that they knew their heroes and therefore to consider the polarization aesthetic.

In either case, the one line which violates these constraints does so, at least in part, on purpose:

Λ 379 ἐκ λόχου ἀμπήδησε καὶ εὐχόμενος ἔπος ηὔδα

This line introduces a speech by a Trojan, and it is not spoken over a dead body. Paris has just shot an arrow into Diomedes’ foot from behind the tomb of Ilos, whereupon he laughs sweetly (i.e. cruelly), leaps out of his ambush and says εὐχόμενος ἔπος:

Λ 380 βέβληαι, οὐδ’ ἅλιον βέλος ἔκφυγεν· ὡς ὄφελόν τοι
          νείατον ἐς κενεῶνα βαλὼν ἐκ θυμὸν ἑλέσθαι.
          οὕτω κεν καὶ Τρῶες ἀνέπνευσαν κακότητος
          οἵ τέ σε πεφρίκασι λέονθ’ ὡς μηκάδες αἶγες

There are two aspects to the constraint-breaking at Λ 379: Diomedes is not dead, and Paris is not Greek. But Diomedes himself points out the first of these abnormalities in his reply to Paris: {90|91}

Λ 385 τόξοτα, λωβητήρ, κέρᾳ ἀγλαέ, παρθενοπῖπα,
          εἰ μὲν δὴ ἀντίβιον σὺν τεύχεσι πειρηθείης
          οὐκ ἄν τοι χραίσμῃσι βιὸς καὶ ταρφέες ἰοί·
          νῦν δέ μ’ ἐπιγράψας ταρσὸν ποδὸς εὔχεαι αὔτως.
          οὐκ ἀλέγω, ὡς εἴ με γυνὴ βάλοι ἢ πάϊς ἄφρων·
          κωφὸν γὰρ βέλος ἀνδρὸς ἀνάλκιδος οὐτιδανοῖο

I am not dead, he says : εὔχεαι αὔτως. Aside from this, he stresses that Paris’ wounding of him with an arrow from ambush is a sign of his effeminate, cowardly fear of manly, competitive single combat until death. Accordingly, it is possible that Diomedes’ reply implies recognition of the breaking of both constraints in Λ 379. For Paris, a Trojan wounding a Greek, to receive the introductory speech formula of a Greek who has slain a Trojan constitutes a formal perversion of the heroic code which parallels precisely his actual perversion of it. But before assuming that the breaking of the Greek/Trojan polarization around (X) is the product of aesthetic and expressive motives, it is necessary to investigate the possibility that it is a compositional slip or the result of compositional stress. In fact some of the formulas for introducing a speech by an εὐχόμενος warrior do not fit Paris’ name-epithet formulas: [


In any case, Paris’ name does not need to be specified at Λ 379, and unless such specification is necessary, the poet avoids it (see below, n. 39). An alternative pair of expressions which do not specify the speaker’s name also exists :


The latter expression will fit perfectly the end of Λ 379:

* ἐκ λόχου άμπήδησε(ν), ἐπευχόμενος δέ προσηύδα

Notice that this Τ2 variant is not isometric with (X), since (X) begins with a consonant and (Y4) with a vowel. Since such a set of options (X, Y3, Y4) is exceptional—metrical ‘licenses’ such as hiatus and lengthening in arsis are permitted at the masculine/feminine caesura—it may be the point of departure for the deeper semantic constraints on (X) and its expressiveness {91|92} at Λ 379. On the other hand, because of these licenses, (Y3) is uneconomical, and it may not be a stable element of the poet’s repertoire. Then the occurrence of (X) at Λ 379 is a product of compositional stress and not expressiveness. The single attestation of (Y4) is ambiguously isolated, but compositional slips are not that rare. However a decision on this provocative problem is not of immediate consequence: we have specified securely other constraints on (X) and (Y) and adduced evidence for the functional equivalence of εὔχομαι in both.

The actual speeches introduced or concluded by these formulas are the next place to look for clarification of the meaning of εὔχομαι. The most common type is a statement that the speaker has killed the warrior whom he stands over:

Ξ 453 Πουλυδάμας δ’ ἔκπαγλον ἐπεύξατο μακρὸν ἀύσας·
          οὐ μὰν αὖτ’ ὀΐω μεγαθύμου Πανθοΐδαο
          χειρὸς ἀπὸ στιβαρῆς ἅλιον πηδῆσαι ἄκοντα
          ἀλλά τις Ἀργείων κόμισε χροΐ, καί μιν ὀΐω
          αὐτῷ σκηπτόμενον κατίμεν δόμον Ἄϊδος εἴσω.
          ὣς ἔφατ’ Ἀργείοισι δ’ ἄχος γένετ’ εὐξαμένοιο.
Ξ 500  πέφραδέ τε Τρώεσσι καὶ εὐχόμενος ἔπος ηὔδα·
          εἰπέμεναί μοι, Τρῶες, ἀγαυοῦ Ἰλιονῆος
          πατρὶ φίλῳ καὶ μητρὶ γοήμεναι ἐν μεγάροισιν·
          οὐδὲ γὰρ ἡ Προμάχοιο δάμαρ Ἀλεγηνορίδαο
          ἀνδρὶ φίλῳ ἐλθόντι γανύσσεται, ὁππότε κεν δὴ
          ἐκ Τροίης σὺν νηυσὶ νεώμεθα κοῦροι Ἀχαιῶν.
          ὡς φάτο, τοὺς δ’ ἄρα πάντας ὑπὸ τρόμος ἔλλαβε γυῖα.
Ν 413 Δηΐφοβος δ’ ἔκπαγλον ἐπεύξατο, μακρὸν ἀύσας·
          οὐ μὰν αὖτ’ ἄτιτος κεῖτ’ Ἄσιος, ἀλλά ἕ φημι
          εἰς Ἄϊδός περ ἰόντα πυλάρταο κρατεροῖο
          γηθήσειν κατὰ θυμόν, ἐπεί ῥά οἱ ὤπασα πομπόν.
          ὣς ἔφατ’ Ἀργείοισι δ’ ἄχος γένετ’ εὐξαμένοιο.

In these passages εὔχομαι no longer designates the statement of a hero’s own birth, but rather the statement of another hero’s death. A simple associative process is responsible for this reversal, as we can see from the following example: [

Φ 182 … Ἀχιλεὺς δ’ ἄρ’ ἐνὶ στήθεσσιν ὀρούσας
          τεύχεά τ’ ἐξενάριξε καὶ εὐχόμενος ἔπος ηὔδα·
          κεῖσ’ οὕτως· χαλεπόν τοι ἐρισθενέος Κρονίωνος
          παισὶν ἐριζέμεναι ποταμοῖό περ ἐκγεγαῶτι·
          φῆσθα σὺ μὲν ποταμοῦ γένος ἔμμεναι εὐρὺ ῥέοντος
          αὐτὰρ ἐγὼ γενεὴν μεγάλου Διὸς εὔχομαι εἶναι.

By killing another hero I assert my own identity and superiority with punctilio. Death-εὔχομαι does not designate or presuppose birth-εὔχομαι {92|93} —instead one suggests the other. In fact, the context of the γενεή-speeches of Glaukos, of Aeneas, and here, of Achilles, is the confrontation of warriors in combat, a confrontation which ideally ends with a death-εὔχομαι speech. [
37] The common feature of both usages of εὔχομαι is proud, contentious, and accurate statement about me. [38] The divergent feature is origin of my life vs. conclusion of your life, but both subjects are appropriate intrinsically to the functionally marked word for ‘say’:

Ζ 206 Ἱππόλοχος δέ μ’ ἔτικτε, καὶ ἐκ τοῦ φημι γενέσθαι·
          πέμπε δέ μ’ ἐς Τροίην, καί μοι μάλα πόλλ’ ἐπέτελλεν,
          αἰὲν ἀριστεύειν καὶ ὑπείροχον ἔμμεναι ἄλλων,
          μηδὲ γένος πατέρων αὐσχυνέμεν, οἳ μέγ’ ἄριστοι
          ἔν τ’ Ἐφύρῃ ἐγένοντο καὶ ἐν Λυκίῃ εὐρείῃ.
          ταύτης τοι γενεῆς τε καὶ αἵματος εὔχομαι εἶναι.

These six lines join together the simple, basic themes with which εὔχομαι is associated in all its secular usages : pride in race, pride in fatherland, and pride in superior achievement. When a hero εὔχεται, he says the most significant facts that he can about himself.

There are, finally, two attestations of εὔχομαι introducing speech which are not formulaic. The first is

Τ 100 ἤτοι ὅ γ’ εὐχόμενος μετέφη πάντεσσι θεοῖσι

which has been discussed above (n. 6). The usage of this expression is exactly parallel to the usage of | εὔχεαι εἶναι #, a formula confined to gods speaking of their sons, and the speech it introduces is Zeus announcement of the birth of Herakles. Accordingly, the usage and meaning of εὐχόμενος here is no less traditional than that of εὔχεαι εἶναι #, and contextual divergence motivates the lexical divergence of the line from the direct-speech εὔχομαι formula (X). Τ 100 is a formula of the type: {93|94}


where the end of the line is usually occupied by the name/name + epithet of the speaker. Here, however, it is occupied by the addressees (which are always plural; thus μετέφη, vs. προσέφη) and the beginning of the line mentions the speaker by the use of a pronoun: [

Τ 100 ἤτοι ὅ γ’ εὐχόμενος μετέφη πάντεσσι θεοῖσι

Otherwise the line exactly parallels

Β 411 τοῖσιν δ’ εὐχόμενος μετέφη κρείων Ἀγαμέμνων

which introduces a speech with prayer structure and in which εὔχομαι has sacral meaning. If the two lines are a rare example of formal parallelism between sacral and secular εὔχομαι formulas, context makes any ambiguity in its meaning flatly impossible since the speaker in Τ 100 is a god, Zeus, not a man. [
40] Gods do not pray unless they are disguised as mortals. Correspondingly, the context of Β 411 is a ritual narrative:

Β 410 βοῦν δὲ περιστήσαντο καὶ οὐλοχύτας ἀνέλοντο

is the line which immediately precedes it. Notice also that εὐχόμενος occupies a slot in the formula which is not lexically fixed, as against other formulas which preserve the sacral/secular split intact. This is a μετέφη-formula, the others are properly εὔχομαι-formulas.

Lastly, there is the line

ξ 463 εὐξάμενός τι ἔπος ἐρέω· οἶνος γὰρ ἀνώγει.

which is formally isolated but analogous to εὐχόμενος | ἔπος in (X). [
41] This suggests the functional equivalence of εὔχομαι in both expressions, and its contextual peculiarities are not such as to necessitate functional divergence either. The first of these, in fact, suggests a reason for its formal {94|95} differentiation from (X): ξ 463 occurs in spoken dialogue, not narrative, and there are signs that it is otiose for a speaker in epic to use the narrator’s formulas for introducing or concluding speech. [42] The second contextual peculiarity is that the speech which follows is not a death announcement, but an adventure tale. We can correlate this with another fact: neither (X) nor (Y) occur at all in the Odyssey, while ξ 463 does. At first this seems natural, since both (X) and (Y) are battle-book formulas and the Odyssey contains no battle-books. But the absence of (X) and (Y) cannot be so easily dismissed. A battle does take place in the Odyssey, and it is described with many of the normal battle-book formulas (see χ 240ff.), including one death announcement:

χ 285 Κτήσιππον δ’ ἄρ’ ἔπειτα βοῶν ἐπιβουκόλος ἀνὴρ
          βεβλήκει πρὸς στῆθος, ἐπευχόμενος δὲ προσηύδα·
          ὦ Πολυθερσεΐδη φιλοκέρτομε, μή ποτε πάμπαν
          εἴκων ἀφραδίῃς μέγα εἰπεῖν, ἀλλὰ θεοῖσι {95|96}
          μῦθον ἐπιτρέψαι, ἐπεὶ ἦ πολὺ φέρτεροί εἰσι
          τοῦτό τοι ἀντὶ ποδὸς ξεινήϊον, ὅν ποτ’ ἔδωκας
          ἀντιθέῳ Ὀδυσῆϊ δόμον κατ’ ἀλητεύοντι.

This Ktesippos is a terrible villain in terms of the morality of the Odyssey, for his throwing a cow’s foot at Odysseus as a ξεινήϊον (υ 296–302) is a vile perversion of the rules of reciprocity whose observance the poem continually stresses. Notice that his murder is not an heroic achievement for Philoitios, but one which he justifies as another ξεινήϊον in return for the cow’s foot, and as a moral lesson for those who ‘talk big’ (μέγα εἰπεῖν). Big talk over dead bodies is actually forbidden in the Odyssey: When Eurycleia sees the suitors dead and wants to express her joy, Odysseus restrains her:

χ 411 ἐν θυμῷ, γρηῦ, χαῖρε, καὶ ἴσχεο μηδ’ ὀλόλυζε·
          οὐχ ὁσίη κταμένοισιν ἐπ’ ἀνδράσιν εὐχετάασθαι.
          τούσδε δὲ μοῖρ’ ἐδάμασσε θεῶν καὶ σχέτλια ἔργα.
          οὔ τινα γὰρ τίεσκον ἐπιχθονίων ἀνθρώπων
          οὐ κακὸν οὐδὲ μὲν ἐσθλόν, ὅτις σφέας εἰσαφίκοιτο·

Such a doctrine – οὐχ ὁσίη κταμένοισιν ἐπ’ ἀνδράσιν εὐχετάασθαι – is completely alien to the Iliad and its battle-books, and the absence of (X) and (Y) from the Odyssey is to be interpreted as genre-suppression, not omission. Thus the elaborate apology which surrounds the only Odyssean occurrence of secular εὔχομαι which is not in a γένος context:

ξ 463 εὐξάμενός τι ἔπος ἐρέω· οἶνος γὰρ ἀνώγει
          ἠλεός, ὅς τ’ ἐφέηκε πολύφρονά περ μάλ’ ἀεῖσαι
          καί θ’ ἁπαλὸν γελάσαι, καὶ τ’ ὀρχήσασθαι ἀνῆκε
          καί τι ἔπος προέηκεν ὅ περ τ’ ἄρρητον ἄμεινον.

The speaker is Odysseus in disguise, about to tell Eumaios and his friends a story about the Trojan war and about Odysseus himself, how he managed to trick Thoas out of his cloak one cold night on an ambush and gave it to his good friend and equal (ξ 470–1), the narrator. [
43] And actually, that {96|97} is the disguised Odysseus’ purpose as well, to get himself a warm cloak on that νύξ σκοτομήνιος. The story is—in more ways than one—proud talk that raises its speaker’s prestige (and almost gives away his identity), but in the Odyssey it receives a moral interpretation as an αἶνος ἀμύμων (ξ 508 by which Odysseus obtains proper treatment as a guest in the form of a χλαῖνα ἀμοιβάς (ξ 520f.)—a symbolic mantle rather than a nice, warm overcoat. This moralism is also responsible for the change in context (and perhaps form as well) of the battle-book formula (X), and it gives us a usage of εὔχομαι which almost has the pejorative connotations of our word ‘boast’. Otherwise, all the attestations in this section are receptive to the meaning ‘say (proudly, contentiously, accurately)’.

b. Section II attestations

The attestations of εὔχομαι in this section do not require close study to determine their meaning. Two of them literally cross-refer to attestations of εὔχομαι in the preceding class (Λ 388 cf. Λ 379; Ξ 484 cf. Ξ 486), four are contextually parallel to the death-εὔχομαι usage (Ν 447, Π 844, Φ 501, Β 597), two are glossed by φημί (Ξ 366, Α 397) and again occur in contexts which are parallel to those of death-εὔχομαι. Formally these eight attestations show no consistency, but that is typical as well of the sacral εὔχομαι expressions which report whole prayers, and the explanation for it is the same (see Ch. II, p. 65f.). Finally there is one attestation which is contextually if not functionally different from the others in this class:

λ 261 ἣ δὴ καὶ Διὸς εὔχετ’ ἐν ἀγκοίνῃσιν ἰαῦσαι
          καί ῥ’ ἔτεκεν δύο παῖδ’, Ἀμφίονά τε Ζῆθόν τε,
          οἳ πρῶτοι Θήβης ἕδος ἔκτισαν ἑπταπύλοιο

This line occurs in the so-called ‘Catalogue of Women’ in the Odyssey, where some of the genealogies are also introduced by φημί:

λ 236 ἣ φάτο Σαλμωνῆος ἀμύμονος ἔκγονος εἶναι
          φῆ δὲ Κρηθῆος γυνὴ ἔμμεναι Αἰολίδαο
λ 306 ἔσιδον, ἣ δὴ φάσκε Ποσειδάωνι μιγῆναι

These are instances of sub-genre language appropriate to catalogue poetry in which φημί, as elsewhere (above, n. 22), is replacing εὔχομαι. λ 261’s particular difference from the (γένος) εὔχομαι εἶναι # formula found elsewhere in epic is the use of an infinitive after εὔχομαι other than εἶναι. This is an aesthetically valuable means of variatio for genealogical catalogues, but the function of εὔχομαι is clearly identical in both the standard epic γένος-formula and in λ 261. [
44] For a semantic parallel to its infinitive which uses εὔχομαι εἶναι #, see Θ 190, above, p. 84.

E. Conclusion


[ back ] 1. On the historical relationship between these pairs and their significance for the history of the dactylic hexameter, see Nagy 1974: Ch. 3, pp. 49–102.

[ back ] 2. On another archaism in the context of this line, Athena’s retreat into the house of Erechtheus, see Nilsson 1921.

[ back ] 3. For archaeological evidence of Cretan hegemony, see Dow 1960. For the meaning of this archaic plural, op. Homeric τόξα, not ‘bows’ but ‘bow and arrows’, and see Schwyzer-Debrunner 1966: p. 43 and n. 3, where a modern Greek parallel (Ikaria: sing. of city, p1. of city + environs) is cited.

[ back ] 4. Nagy 1974: Ch. 3, Ch. 4, pp. 49–117. The initial observations leading up to Nagy’s reconstruction were made by Watkins 1969, p. 227.

[ back ] 5. Formulas of pherecratic shape are the basic wherewithal of epic diction, as Nagy 1974 has shown. This lyric meter with triple dactylic expansion is in fact the ancestor of the hexameter according to his hypothesis.

[ back ] 6. The first two are simple transformations which need no comment (see the complete list), but the last is peculiar. The transformation to second person singular is accompanied by a reversal in hierarchy. Normally a child εὔχεται concerning his father, while in HHerm 378 and ι 529 the fathers εὔχονται concerning their sons. This phenomenon is restricted to the male children of male gods, Polyphemos son of Poseidon and Hermes son of Zeus. In one place in the Iliad we have what a god actually εὔχεται about the birth of his son. Not surprisingly, it is Zeus speaking of the imminent birth of Herakles, T 100ff. When the son himself is talking, he uses the second person; but the phenomenon recurs with the same contextual restriction when another speaks of the genealogy of a male god’s son, ι 519 (Polyphemos) and, with φημί substituting for εὔχομαι (on this usage, see below, pp. 76ff.),

          O 112 Ἀσκάλαφος, τόν φησιν ὃν ἔμμεναι ὄβριμος Ἄρης

The use of the second person singular by these individuals is to be correlated with their highly prestigious birth: it gives them the privilege to se tutoyer their divine fathers and their fathers the pride to εὔχεσθαι about them. On similar naming constraints, see below, n. 9, n. 11, and, on the second person singular, n. 16.

[ back ] 7. Parry 1928a: pp. 83–5, pp. 206–7 = A. Parry 1971: pp. 66–7, p. 164 excepts examples of this sort from his rule that ornamental epithets separated from their nouns carry real meaning. For a valuable critique of this rule itself, see Hainsworth 1968: pp. 90–92 if. Note that the special problems of ornamental epithets do not complicate our understanding of expressions like γένος εὔχομαι εἶναι. The epithets provide an analogy weighted against the argument we are making.

[ back ] 8. One is reminded of the mythological motif of autochthony, and also of the expression πατρίδα γαῖαν κ.τ.λ. (Consider Γαῖα Earth-mother, in mythology.) See also the expression τίς πόθεν εἰς ἀνδρῶν; (Odyssey, passim), discussed below, pp. 123ff.

[ back ] 9. The etiquette rules which restrict the information given in these lines are a sign that we are dealing with name-conventions. Description of these rules is not relevant to our immediate purposes, but one notable example is worth mentioning since it precisely obscures the fact that these are naming-statements: it is unconventional to mention one’s own name. In the γενεή-speeches of Glaukos, Asteropaios, and Aeneas, however long they are, the heroes never do so. The same is true in the formulas under discussion. Moreover, we can see this constraint in operation by comparing the two genealogical doublets of Aeneas:

          E 247 Αἰνείας δ’ υἱὸς μὲν ἀμύμονος Ἀγχίσαο

                   εὔχεται ἐκγεγάμεν, μήτηρ δέ οἵ ἐστ’ Ἀφροδίτη.

Here Sthenelos is telling Diomedes that Aeneas is approaching. But when Aeneas identifies himself to Achilles, he must omit his own name:

          Y 208 αὐτὰρ ἐγὼν υἱὸς μεγαλήτορος Ἀγχίσαο

                    εὔχομαι ἐκγεγάμεν, μήτηρ δέ μοί ἐστ’ Ἀφροδίτη·

In three places, however, the convention is broken.

          ι 19 εἴμ’ Ὀδυσεὺς Λαερτιάδης, ὃς πᾶσι δόλοισιν

                 ἀνθρώποισι μέλω, καί μευ κλέος οὐρανὸν ἵκει.

          Α 180 Μέντης Ἀγχιάλοιο δαΐφρονος εὔχομαι εἶναι # υἱός

          HApoll 480 εἰμὶ δ’ ἐγὼ Διὸς υἱός, Ἀπόλλων δ’ εὔχομαι εἶναι

In two of these, Apollo and Odysseus are identifying themselves after emerging from a disguise. The third is Athena-Mentes, who is in disguise, a fact which does not seem to fit. But the disguise-motif is relevant as a secondary phenomenon (see below). Basically, the unconventional statement of one’s own name is a sign of special fame. It is, in fact, comparable to referring to one’s self in the third person, for in the third person mention of a person’s proper name is normal (of. E 247, a 418, E 246, N 54, etc.). But just as people who use the third person singular of themselves are asserting their fame (ep. Napoleon in his memoirs, Caesar in his commentaries, Achilles in A 240), so Homeric characters who mention their own names are implying that they are famous. Odysseus, our ‘hero’, even makes it explicit: (ι 20) καί μευ κλέος οὐρανόν ἵκει (note further that he has adopted the persona of the narrator, the third person par excellence, at this point in the poem). In Apollo’s case, his status as a god who is also the ‘hero’ of the hymn and his emergence from disguise motivate the anomalous usage, and in Athena’s, it is perhaps to offer a counter-example to Telemachus’ unheroic and demoralizing doubt about who his father is (α 215–6, see below, p. 77 f.) and therefore who he himself is that she takes special pride in her own name (cp. α 321 ὑπέμνησέν τέ ἑ πατρός with the narrator punning on her name Μέντης). Her disguise is, from this point of view, pregnant irony.

[ back ] 10. On the phraseological change in the first line of Υ 208 which derives from the third to first person grammatical change, see the preceding note.

[ back ] 11. On this convention, see West 1966: p. 431 ad Theog. 1002. For Achilles’ genealogy, see Π 860.

[ back ] 12. The reason it does not recur maybe banal, —i.e. simply because the idea does not surface again but it is tempting to see aesthetic purification of the diction at work in its suppression. Note the repetition of ἔμμεναι and the slightly forced use of αὐτάρ, which normally introduces a new grammatical subject.

[ back ] 13. For the idea, compare

          Y 105 καὶ δὲ σέ φασι Διὸς κούρης Ἀφροδίτης

                   ἐκγεγάμεν, κεῖνος δὲ χερείονος ἐκ θεοῦ ἐστίν·

                   ἣ μὲν γὰρ Διός ἐσθ’, ἣ δ’ ἐξ ἁλίοιο γέροντος.

Here φασι is a simple substitute for εὔχομαι.

[ back ] 14. See Benveniste 1966: pp. 225–236 for a lucid and aesthetic discussion of this aspect of the third person as against the first. On the constraint against second person (twice only), see above, n. 6. N. 16 below is also relevant.

[ back ] 15. This semantic contrast, which is consistent in Epic, is confused by LSJ9 s.vv. and not noticed elsewhere, to my knowledge. In later Greek, it takes a similar but horizontal form, see Herod. Ι. 35.1 Φρὺξ μὲν γενεῇ, γένεος δὲ τοῦ βασιληίου, where γένος = family, γενεή = ethnicity.

[ back ] 16. It is possible, though perhaps overly subtle, to interpret the use of φημί in the second person middle as against φασί in Υ 206 as a concession to this artificial kind of rank-pulling and to the relative similarity of their immediate ancestry. Asteropaios has enough prestige so that Achilles can se tutoyer with him. See n. 6 above.

[ back ] 17. Contrast Odysseus in disguise as a nobleman who has fallen upon hard times. Although or, more accurately, because he is outwardly disgusting and in fact a bastard, he uses εὔχομαι twice of his γένος (ξ 199 [place], ξ 204 [father]). See also n. 9, fin. on Athena’s response to Telemachus’ unheroic attitude.

[ back ] 18. See above, n. 9.

[ back ] 19. See the discussion of this line above, pp. 50ff., below, n. 23, as well as Ch. II n. 67, n. 68. See also Κ 50 for a contrasting view of his birth, and in a similar vein compare the proper format for divine genealogy in n. 6 above. My final views on this problematic line are expressed in the Epilogue, below p.117 and nn. 28 and n. 46, where I refer to a forthcoming discussion of Hector’s mythical functions by Frame.

[ back ] 20. For other examples of the functional relationship between εὔχομαι and φημί, see the subsequent discussion and the following list of additional examples (a partial list accumulated in reading the poems, since the concordances do not list φημί, εἶναι, ἔμμεναι):

          σ 218 καί κέν τις φαίη γόνον ἔμμεναι ὀλβίου ἀνδρὸς

          O 112 Ἀσκάλαφος, τόν φησιν ὃν ἔμμεναι ὄβριμος Ἄρης

          HAphr 284 φασίν τοι νύμφης καλυκώπιδος ἔκγονον εἶναι

          Y 105-106 καὶ δὲ σέ φασι Διὸς κούρης Ἀφροδίτης # ἐκγεγάμεν

          Z 100 ὅν πέρ φασι θεᾶς ἐξέμμεναι·

          Z 206-211 Ἱππόλοχος δέ μ’ ἔτικτε, καὶ ἐκ τοῦ φημι γενέσθαι·

                           … ταύτης τοι γενεῆς τε καὶ αἵματος εὔχομαι εἶναι

          Φ 159                               … τὸν δ’ ἐμέ φασι # γείνασθαι·

          λ 236 ἣ φάτο Σαλμωνῆος ἀμύμονος ἔκγονος εἶναι

[ back ] 21. Hainsworth 1968: pp. 137–8. Many more expressions of this shape are not so moved because of Hermann’s bridge. Movement is possible for a total of 56 expressions.

[ back ] 22. This fixation is due to constraints of inner metric and syntactic problems in its transferral. Contrast the mobility and apparent phraseological flexibility of φημί – γένος expressions. It is likely that φημί is in fact replacing εὔχομαι in this usage. (For a parallel, cp. the replacement of κλύω by ἀκούω, above, Ch. II, n. 57).

[ back ] 23. In light of this explanation, the extensive anomalies of Ν 54:

          Ἕκτωρ ὃς Διὸς εὔχετ’ ἐρισθενέος πάις εἶναι

are increased by its unparalleled separation of εὔχομαι from εἶναι in a γένος context. See n. 19.

[ back ] 24. For these terms and concepts, see Ch. II, above, p. 46 and n. 62.

[ back ] 25. Nagler 1967: pp. 269–311.

[ back ] 26. Part of the brilliance in this passage is that Agamemnon’s words are not completely clear: does he mean death for Achilles? or does he foresee death for himself and his army if Achilles refuses the ransom?

[ back ] 26a. I suggest that the covert reference here is to the war-god Ares himself. The simplex epithet is his exclusive epithet among gods and men (Ε 704, etc.), being otherwise applied to weapons only (ἔγχος, αἰχμή, etc.). Likewise, παγχάλκος/-εος applies here to θεός elsewhere to ἄορ, ῥόπαλον and κυνέη only (θ 403, λ 557, χ 102, and σ 378). Finally, see Batrach 130.

[ back ] 26b. See n. 27 on a forthcoming paper by G. Nagy.

[ back ] 27. We may justly wonder if Agamemnon is ‘accurately’ ἄριστος, but in terms of the concrete criteria he himself mentions— βασιλεύτερος, προγενέστερος —it cannot be denied. It is, after all, only Achilles who questions and demolishes the accuracy of this social label ἄριστος, which is the exclusive property of the chief hero in his own epic (Achilles in the Iliad, Odysseus in the Odyssey) but not in each other’s epic : see a study of ἄριστος in a forthcoming work by G. Nagy entitled “Demodocus, Odyssey, Iliad” [see now G. Nagy, Best of the Achaeans, 2nd. ed. 1999].

[ back ] 28. A possible etymology of πόσις is relevant in this discussion : “Die Bed. ‘Herr, Gatte’ wird allgemein aus einer älteren Bed. ‘selbst’ in lit. pàts … ebenso wie in der Identitäts- und Verstärkungspartikel lit. pàt, heth. -pat (-pit, -pe) ‘eben(so), ebenfalls, gerade’ erklärt” (Frisks s.v. πόσις). This etymology, however, is severely criticized by Szemerényi 1964: pp. 337ff.; contra Benveniste 1969: t. 1, pp. 87–91 and 1966: pp. 301ff. See also below, pp. 119–21, for a Vedic parallel to this expression.

[ back ] 29. The word is used three times of Leodes (φ 145, χ 318, χ 321), who is the first suitor to try and fail at the bow of Odysseus, and once as a substantive in Ω 221 μάντιες … θυοσκόοι ἢ ἱερῆες. It is very probable that this archaic word is attached to a sacral function as important as that of μάντις or ἱερεύς, but the Homeric contexts are not complete or informative enough to make a precise judgment. That Leodes is an ignominious figure and that θυοσκόοι in Ω 221 is perceptibly contemptuous does not alter this probability, since contempt for seers is a cliché. However, the cliché may be to mind in χ 321, in which case θυοσκόος εὔχεαι # is as sarcastic as μετὰ τοῖσι … εὔχεαι εἶναι #. See the subsequent discussion, and below, p. 119.

[ back ] 30. For identification on the basis of occupation, compare the banal Smith, Cooper, etc.

[ back ] 31. Benveniste 1969: t. 2, pp. 252–4.

[ back ] 32. The consequences of crossing Zeus ξείνιος are not made explicit in that passage, but elsewhere they are violent, massive destruction (e.g. Ν 624–5). I intend to discuss the sanction in a forthcoming study of the word μῆνις.

[ back ] 33. Mauss 1967 : pp. 10–1 with copious references to other societies which I have omitted.

[ back ] 34. It is worth mentioning that both ξεῖνος and ἱκέτης occur as onomastics from the Mycenaean period on (ke-se-nu-wo PY Cn 286, MN ; i-ke-ta KN 799, MN) (later Ἱκετάων, Ἱκέτυλλος; Πρόξενος, Ξενοφῶν, κ.τ.λ). Also the reader should note that the element of ‘contentiousness’ in ξεῖνος/ἱκέτης + εὔχομαι is not directed toward the partner in the relationship, but it is an aspect of the pride of the ἱκέτης/ξεῖνος which is in principle directed at those who are not related to the person specified. I say ‘in principle’ if not ‘in fact’, because the significant aspect of these qualifiers (‘accurately’, ‘contentiously’, etc.) is that they are dimensions of the marked function of εὔχομαι as against φημί. The rigid fixation of its meaning down to the details of those dimensions is an aspect of formulaic, poetic language which would be remarkable in a natural language. It is not impossible therefore that the rigidity is at some points softened by the tradition, which after all consists of people who also spoke a natural language. But this occurs very rarely, and not, in my opinion, here.

[ back ] 35. Πάρις ( ) and once Πάρῑς ( –); || Ἀλέξανδρος θεοειδής#; # δῖος Ἀλέξανδρος || ; Ἀλέξανδρος || or || Ἀλέξανδρος.

[ back ] 36. For the two notions literally juxtaposed in a death-εὔχομαι speech, cp. Υ 390 ἐνθάδε τοι θάνατος, γενεὴ δέ τοί ἐστ’ ἐπὶ λίμνῃ

[ back ] 37. This explains why other speeches introduced and/or concluded by (X) or (Y) are not statements that the speaker has killed the person he is standing over, but challenges to confrontation with another hero (e.g. Υ 424ff.). Confrontation ideally begins with γένος-speeches, which are in fact challenges (explicitly : Υ 200 – 9), continues with actual combat, and concludes with death-εὔχομαι speeches. But in cases such as Υ 424ff. there is dovetailing of one confrontation with the next which substitutes a challenge for the death announcement; at Ν 445ff. there is again dovetailing, where a death-εὔχομαι speech formula introduces a γένος-speech. (For the term dovetailing, as it applies to metrical phenomena, see Maas 1962: p. 44 and Nagy 1974: pp. 279–302.)

[ back ] 38. Thus also Paris (Λ 379) can try to εὔχεσθαι that he has wounded Diomedes. It is a misuse of the term, since he has wounded him only and wounded him with bow and arrow, and this is not a definitive, prideful achievement in terms of the strict heroic code. Ironically, however, it is a characteristic achievement for Paris, whose motto νίκη δ’ ἐπαμείβεται ἄνδρας (Ζ 339) is not simply unheroic but anti-heroic, since it denies the prowess of the individual.

[ back ] 39. For the significance of introductory speech formulas which identify and do not identify the speaker, see above, p. 21 and Parry 1937 : pp. 59–63 = A. Parry 1971: pp. 414–8.

[ back ] 40. See note 39. The differences between the two have no relationship to the sacral/secular split in εὔχομαι. For the single instance of (X) in a sacral context at Κ 461, see the discussion p. 133f.

[ back ] 41. For evidence that εὐχόμενος and ἔπος are traditionally and syntactically bound together (ἔπος being cognate accusative), see below, Etymological Epilogue, p. 127ff. This is true in spite of the separate existence of ἔπος ηὔδα#, as in

          Ζ 54                                || καὶ ὁμοκλήσᾱς ἔπος ηὔδα #

          Κ 377                               || ὁ δέ δακρύσᾱς ἔπος ηὔδα #

          Μ 163                              || καὶ ἀλαστήσᾱς ἔπος ηὔδα #

          Ω 307                              || καὶ φωνήσᾱς ἔπος ηὔδα #

          ν 199                               || ὀλοφυρόμενος δ’ ἔπος ηὔδα

          Ρ 119                               || παριστάμενος ἔπος ηὔδα


          (X) 6 times                       || και εὐχόμενος ἔπος ηὔδα

Notice that only one (Ρ 119) of the examples of ἔπος ηὔδα # shows awareness of the digamma in Ϝέπος (the others have surallongement if the digamma was felt), while (X) cannot scan without it. From a historical point of view, the other ἔπος ηὔδα # expressions are secondary to (X)’s εὐχόμενος ἔπος. See also Simonides 543.25 – 6 (Page), ὅττι δὲ θαρσαλέo̅v ἔπος εὔχομαι/ἢ νόσφι δίκας, where εὔχομαι again has secular meaning, and C. Watkins (forthcoming article).

[ back ] 42. For another example of formal divergence between phrases which are functionally parallel but occur in spoken dialogue as against narrative, cp.

          Ψ 149 ὣς ἠρᾶθ’ ὁ γέρων, σὺ δέ οἱ νόον οὐκ ἐτέλεσσας

which concludes a prayer in dialogue. Transformations of (A) exist with negative responses to match the function of Ψ 149, but this expression is parallel to a secular series:

          Β 330 κεῖνος θ’ ὣς ἀγόρευε· τὰ δὴ νῦν πάντα τελεῖται

          σ 271, Ξ 48 κεῖνος τὼς ἀγόρευε· τὰ δὴ νῦν πάντα τελεῖται

          θ 570 ὧς ἀγόρευ’ ὁ γέρων· τὰ δέ κεν θεὸς ἢ τελέσειεν

          ν 178 ὧς ἀγόρευ’ ὁ γέρων· τὰ δὲ δὴ νῦν πάντα τελεῖται

All of these occur only in spoken dialogue. Another transformation actually occurs in narrative, where contextual anomalies rule out (A) or any of its transformations:

          γ 62 ὣς ἄρ’ ἔπειτ’ ἠρᾶτο καὶ αὐτὴ πάντα τελεύτα

Athena in disguise has just made a prayer to Poseidon, and for another god to hear Athena’s prayer (as the cadence of (A) in all its transformations necessitates) would not suit her dignity. This formula is used so that she can fulfill her own prayer. Perhaps I need not mention that the exception to the etiquette rule I am hypothesizing is Odysseus, who uses the narrator’s speech formulas all the time in ι-μ. This only proves that he has the narrator’s persona.

[ back ] 43. It is interesting to note that in the Doloneia, a genre piece with many themes and even many details (weather, time, etc.) in common with this story, it is Odysseus (not some ragged companion) alone of the five Greek heroes who fails to wear a cloak (χλαῖνα), but carries only his shield (σάκος, Κ 149, vs. cloaks of other heroes elaborately described: Κ 23–4, 29–30, 133–4, 177–8). Compare ξ 482:

          ἀλλ’ ἑπόμην σάκος οἶον ἔχων καὶ ζῶμα φαεινόν

The coincidence is all the more striking, since there is no point made of his lack of a cloak in the Iliad. Is epic too unsophisticated for a cross-reference here in the Odyssey to another version of this tale in which Odysseus himself needed a χλαῖνα – a cross-reference which, like the use of εὐξάμενος, hints broadly that the cloak-less narrator is Odysseus himself? In any case the stories must belong to the same ‘song’ as Lord uses the word (see Lord 1960: pp. 99–123, and esp. p. 102ff. for Yugoslav parallels to the inconsequential detail about Odysseus in the Iliad tale, a typical feature of oral performance).

[ back ] 44. In any case, φημί and εὔχομαι in these attestations are not “perfunctory verbs of speaking” as Denys Page (Page 1955: p. 50, n. 19, and p. 36) calls them, nor are they signs that “the poet sometimes remembers, sometimes forgets that what (the women) say is addressed to Odysseus, not by himself to the audience”. (In several genealogies in the catalogue, neither φημί nor εὔχομαι occur, e.g. λ 267ff., λ 299ff.). εὔχομαι and φημί in this usage are an epic naming convention and never narrate direct speech (i.e. , they are never used as our “…,” she said, etc.). In fact, they are addressed to no one in particular. Page believes that the catalogue is a later insertion and wants its poet to be a clumsy adaptor, which he is not. For an extremely interesting parallel to this usage which also uses φημί to address no one, see the recently (1957) published papyrus fragment of a Hesiodic poem about (appropriately enough) the descent to Hades of Peirithoos and Theseus (Merkelbach and West 1967: pp. 139f., frg. 280). The passage preserved is a dialogue between Meleager and Theseus in the underworld, and in the speech of Theseus to Meleager (introduced by the verb καταλέξω) forms of φημί occur twice in mutilated contexts (11.13, 15) which are perceptibly genealogical and also twice in places which are convincingly restored (11.18-22):

                                                  ἐγγυτέρω γάρ

          φήσ’ εἶ]ναι γεγαὼς αὐτὸς μεγάλου Ἀΐδαο

          Φερσεφ]όνηι κούρηι Δημήτερος ἠυκόμοιο· 20

          αὐτὸς] μὲν γάρ φησι κασίγνητος και ὄπατρος

           ….. .] εν]] Ἀΐδην δὲ φίλλον πάτρωα τε̣τ̣ύχθαι·

                    18–19 Latte

                    22 τῆς ἔμμεναι Merkelbach

          “He (Peirithoos) says he is more closely related by birth than great

          Hades to Persephone, fair-haired Demeter’s daughter;

          for he says he (is her) brother by the same father and that

          Hades became his beloved uncle …”

Theseus is not reporting a conversation he had with Peirithoos but reporting his genealogy; note also the contextual (descent to underworld) parallelism of this catalogue with the Odyssey’s, and its parallel use of the sub-genre formula with φημί.

[ back ] 45. Thus, for example, we have no evidence to favor Benveniste’s suggestion (Benveniste 1969: t. 2, p. 240) that secular εὔχομαι is a variety of sacral εὔχομαι in which the speaker consecrates himself or vows himself metaphorically to be the son of X or to be the most brave. (See further below, p. 104.) This meaning ‘vows himself’ exists in Latin (cp. the archaic custom of devotio discussed by Benveniste ibid.: p. 237) but not, in my opinion, in Homeric Greek, and Benveniste’s attribution of it to secular εὔχομαι is not a rigorous application of the comparative method. Cp. the dictum of Franz Skutsch (cited by Palmer 1963: p. 187) “Look for Latin etymologies on the Tiber”.