Stephanie W. Jamison, H. Craig Melchert, and Brent Vine (eds.). 2013. Proceedings of the 24th Annual UCLA Indo-European Conference. Bremen: Hempen. 87–101.

The Hymnic Long Alpha: Μούσας ἀείδω and Related Incipits in Archaic Greek Poetry*

JOSHUA T. KATZ

Princeton University

Homer’s Iliad, conventionally considered to be our earliest Greek poem,1 begins, of course, as follows:2

##Μῆνιν ἄειδε, θεά, Πηληϊάδεω Ἀχιλῆος / οὐλομένην Of the destructive wrath of Achilles son of Peleus sing, o goddess.

But there are also two non-canonical incipits, the details of which can be found in the apparatus of Martin L. West’s Teubner edition of the Iliad (1998:3) as well as in §B1 of the “Appendix Romana” in his Loeb of Homerica (2003:454–7).3 I cite here the text and translation from the latter:4

Ἡ δοκοῦσα ἀρχαία Ἰλιάς, λεγομένη δὲ Ἀπελλικῶντος [Schimberg (-ῶνος Ribbeck5): ἀπ’ ελικῶνος cod.], προοίμιον ἔχει τοῦτο·

Μούσας ἀείδω καὶ Ἀπόλλωνα κλυτότοξον,

ὡς καὶ Νικάνωρ μέμνηται καὶ Κράτης ἐν τοῖς Διορθωτικοῖς. Ἀριστόξενος δὲ ἐν α΄ Πραξιδαμαντείων (fr. 91a Wehrli) φησὶ κατά τινας ἔχειν·

* I am grateful to the editors for giving a home to this paper, which extends a small section of

the talk (“Gods and Vowels”) that I delivered in October 2012 at the 24th Annual UCLA Indo-European Conference. A full written version of the talk is in press in another volume of pro-ceedings (Katz forthcoming); occasional phrases from the longer paper are repeated here. Al-though, sadly, we never had the chance to talk about these ideas, my expressive thanks go to Calvert Watkins (13 March 1933–20 March 2013).

1 Conventionally, but far from uncontroversially: see, e.g., West 2012 (synthesizing nearly half a century of thoughts on the subject).

2 Throughout, #X means that X is verse-initial; X# that it is verse-final; ##X that it comes at the very beginning of a poem; and X## that it comes at the very end.

3 West (2003:450 n.45) writes of the Appendix Romana, also known as Life of Homer 10, that “[b]y this title I designate further material from the prolegomena to the D scholia on the Iliad in cod. Rom. Bibl. Nat. gr. 6, the manuscript from which the first Anonymous Life is taken”; see also West 2001:73.

4 See also now Nagy 2010b:110, whose version is slightly different. 5 See also West 2001:73, with n.86. I note that in the earlier Teubner edition West prefers

Ἀπελλικῶνος.

Joshua T. Katz 88

Ἔσπετε νῦν μοι, Μοῦσαι Ὀλύμπια δώματ’ ἔχουσαι, ὅππως δὴ μῆνίς τε χόλος θ’ ἕλε Πηλείωνα Λητοῦς τ’ ἀγλαὸν υἱόν· ὃ γὰρ βασιλῆϊ χολωθείς.

What is considered the old Iliad, the one known as Apellicon’s, has this proem:

Of the Muses I sing, and Apollo of the famed bow,

as recorded both by Nicanor and by Crates in his Text-critical Notes. Aristox-enus in book 1 of his Praxidamanteia says that according to some it had:

Tell me now, Muses who dwell on Olympus, how it was that wrath and anger seized Achilles and Leto’s glorious son; for he, angry with the king

(etc.).

The three verses attributed to the Peripatetic philosopher Aristoxenus of Ta-rentum will not occupy me here.6 But what of Apellicon’s evidently hymnic opener, Μούσας ἀείδω καὶ Ἀπόλλωνα κλυτότοξον? Apellicon of Teos, as West points out in a note (2003:455 n.48), was a book collector who died in 84 B.C. and whose library Sulla brought to Rome. In this library there seems to have been an ἀρχαία Ἰλιάς that began differently, in a way that the philosopher, grammari-an, and Homeric critic Crates of Mallos before him (2nd cent. B.C.) and the grammarian Nicanor “Stigmatias” after (early 2nd cent. A.D.) knew and recorded as well.7 West calls Apellicon’s incipit “peculiar.” But is it?

6 West (2011:81) provides a (negative) assessment of their quality, and see also now West

2013:22. 7 West (2001:73) sorts out the details as follows:

The citation of ‘Nicanor and Crates’ [in the Appendix Romana] probably means ‘Nicanor, who mentions Crates as his source’. Crates, however, could not, for chronological reasons, have described the book as Apellicon’s. Apellicon was a rich collector from Teos who acquired from the descendants of Neleus of Scepsis the di-lapidated library of Theophrastus, which included some manuscripts of Aristotle. On Apellicon’s death in 84 BC Sulla brought his library from Athens to Rome; there it was rummaged over by the grammarian Tyrannion, whose work on accentuation is often cited by Herodian. If Apellicon’s name is rightly recognized in our text, we may speculate that Tyrannion found among Apellicon’s books an Iliad with the pe-culiar incipit that he had read of in Crates’ Diorthotika—very likely the same vol-ume—and that Nicanor got his information from Tyrannion.

The Hymnic Long Alpha 89

The answer to this question depends on the point of view one adopts. There are at least four, all of them in principle reasonable. First, the Iliad, including its opening verses, can be—and for that matter nearly always is—read synchronical-ly, as a codified text (whether West’s or someone else’s). Second, it can be con-sidered in its performance context, which might well mean allowing for a prefatory hymn (and presumably for other variations from a codified text as well). Third, it can be seen as a series of possible texts, as something like what Gregory Nagy (in repeated attacks on West and his “Urtext”) calls “multitexts,” which owe their existence to a complicated dance between synchrony and rela-tively shallow diachrony: the dynamics of oral tradition and transmission coupled with the lack of a unitary mode of Homeric performance.8 And fourth, it can be understood as an example of traditional poetry and examined purely diachronical-ly, with an eye to uncovering its deeper (pre-Greek, possibly Proto-Indo-Euro-pean) past.

Only the first is widely viewed as uncontroversial: from our perspective today—and, to judge from the testimonia that have come down to us, from the perspective of most ancients—Apellicon’s incipit certainly is peculiar, being a deviation from the norm that no modern edition prints except, sometimes, in the fine print of the apparatus. On the other hand, it is well known, as West (2011:81) puts it himself, that “[e]pic recitations were commonly prefaced by a hymn to a deity, of the kind represented by the extant Homeric Hymns,”9 and so ##Μούσας ἀείδω καὶ Ἀπόλλωνα κλυτότοξον is definitely not at all peculiar to someone like Nagy, whose battle to convince colleagues that there are “as many Homers as you please” (Pelliccia 1997) appears to be reaching a climax with the “vital online tool” known as “The Homer Multitext Project,” on which he has been working with a “multigenerational team” of acolytes.10 As far as the ἀρχαία Ἰλιάς and its hymnic proem are concerned, Nagy in his recent monograph Homer the Pre-classic (2010b) presses their inner-Greek antiquity very hard vis-à-vis what he calls “the newer Iliad” (or “the Homeric Iliad”) and its familiar “acephalic

Crates was associated with the Library of Pergamum, and Nagy (1998:215–6 and 2010b:110),

who in these and other publications states as fact that he was the Library’s “head,” proposes that Apellicon’s “Homerus auctus” came from there.

8 See among much else the first half of Nagy 2004. 9 West continues: “The practice is alluded to at θ 499 when Demodocus θεοῦ ἤρχετο; cf. also

Pind. Nem. 2.1–3”; see also, e.g., Ford 1992:23–31. 10 See Nagy 2010a; see also http://www.homermultitext.org/about.html and http://homermultitext

.blogspot.com.

Joshua T. Katz 90

prooimion” (##Μῆνιν ἄειδε, θεά, …), referring to the non-standard version as Orphic and comparing Homeric Hymn 25, which begins ##Μουσάων ἄρχωμαι Ἀπόλλωνός τε Διός τε “From the Muses let me begin, and Apollo and Zeus.”11 I do not have the space to rush into the fray and shall say only that it is possible to appreciate the existence of and possible reasons for textual fluidity in the Iliad, a poem in which variae lectiones of a dramatic sort are not abundant, without mak-ing grandiose pronouncements about what the variants might mean on a historical or text-critical level. It is an unfortunate fact that there are significant gaps in our evidence, even after the Archaic age, and West (2011:81) is correct to note simp-ly that “we do not know whether the Panathenaic recitations [of the Iliad and for that matter the Odyssey] began with a hymn (such as H. Hom. 28 [to Athena]).”

This leaves the fourth perspective, the deep historical one. While looking at Archaic epic in this way might seem to jibe better with Nagy’s ideas than with the usual view of a more or less invariant Homer, this is not really so: compatible with both and not a replacement for either, diachronic analysis adds a dimension to the synchronic study of any text, whether fixed or multiform, and it is perhaps most useful to the standard critic in providing explanations for things that do not appear to make sense. As I have put it elsewhere (Katz 2010:360 and 2013:173), the mantra of the historical linguist is this: if you encounter a synchronic oddity, search for a diachronic explanation.

Let us return to ##Μούσας ἀείδω καὶ Ἀπόλλωνα κλυτότοξον, of which West (2011:81) writes that Crates knew it “presumably [from] a rhapsode’s text.” The big-picture oddity has of course to do with the status of this verse, and in particu-lar with its antiquity at the head of something that we might want to call the Iliad. I do not myself wish to argue that it is old in this function. Nevertheless, it may well be old as such. It is clear, after all, that hymnic proems play a significant role in Archaic Greek poetry: both of the major Hesiodic works have one (Theogony 1–115 and Works and Days 1–10), though Crates athetized them, as he presum-ably also did for the ἀρχαία Ἰλιάς;12 these proems resemble the (shorter) Homeric Hymns; and there is linguistic, metrical, and general cultural continuity—though the details are very hard to pin down—between the genre of early Greek epic and

11 See Nagy 2010b:109–19. With the term “acephalic prooimion” Nagy “refer[s] to an epic pro-

oimion that has no naming of the hymnic subject” (119). 12 See above all West 1966:150 and also 1978:65–6 and 136–7. For Gregory Nagy’s view of

Crates, Apellicon’s incipit, and ancient recensions of Homer, see Nagy 1998:213–30 and pas-sim.

The Hymnic Long Alpha 91

its pre-Greek hymnic predecessor (see, e.g., Katz 2005), a style whose language and diction led in other branches of the Indo-European family to such repositories of archaic material as the Indic Rigveda and the Avestan Gāthās. In any event, there is also a small-picture oddity with Apellicon’s incipit, namely that the alpha of ἀείδω scans long. This, I believe, is more interesting than is regularly assumed, and as we shall see, an examination of the linguistic background of ##Μούσας ἀείδω suggests that the little oddity fits together with the big one.

In a forthcoming paper titled “Gods and Vowels,” I point out a number of neglected features that a historical/comparative linguistic examination of our ear-liest Greek incipits reveals.13 For example, if we accept that μῆνις ‘wrath’ and Μοῦσα ‘Muse’ both go back ultimately to *men- ‘think (vel sim.)’ or its deriva-tive *mneh2- ‘keep in mind’, then it is immediately striking, from a diachronic perspective, that the usual start of Homer’s Iliad and Odyssey and the hymnic proems of Hesiod’s Theogony and Works and Days all reflect this root of mental force:

##Μῆνιν ἄειδε, θεά, … (Iliad 1.1) “Of the wrath sing, o goddess.” ##Ἄνδρα μοι ἔννεπε, Μοῦσα, … (Odyssey 1.1) “Of the man sing to me, o Muse.” ##Μουσάων Ἑλικωνιάδων ἀρχώμεθ ̓ ἀείδειν (Theogony 1) “From the Heliconian

Muses let us begin to sing.” ##Μοῦσαι Πιερίηθεν, ἀοιδῇσι κλείουσαι (Works and Days 1) “O Muses from Pieria,

glorifying in songs.”

Furthermore, three of these four (indeed, the ones whose very first syllable goes back to the root) couple a reflex of *men-/mneh2- with one of *h2u̯eid- ‘sing’ (> ἀε/οιδ-).14 And if we move to the Homeric Hymns, the pattern continues, for three of the four major ones, as well as a number of the others, both begin and end with forms of *men-/mneh2- and/or *h2u̯eid-, a pair of roots that between them capture the two main functions of the Indo-European bard, namely to re-member and to sing (note the common closing tag μνήσομ’ ἀοιδῆς##):

ἀείδειν# — μνήσομ’ ἀοιδῆς## (Homeric Hymn [2] to Demeter 1 — 495) [I begin] to sing — I shall remember a song.

13 I hope that the paper will appear soon and thus forbear repeating references to secondary lit-

erature or many of the observations I make there. 14 The presence of the imperative ἔννεπε rather than ἀειδ- in the first verse of the Odyssey may

not be more than mildly bothersome: note the appearance of both the nominal form ἀοιδῇσι and the imperative ἐννέπετε (2) at the start of the Works and Days.

Joshua T. Katz 92

##Μνήσομαι — μνήσομ’ ἀοιδῆς## (Homeric Hymn [3] to Apollo 1 — 546) Let me remember — I shall remember a song.

##Ἑρμῆν ὕμνει, Μοῦσα — μνήσομ’ ἀοιδῆς## (Homeric Hymn [4] to Hermes 1 — 580)

Of Hermes sing, o Muse — I shall remember a song.15

Of immediate interest is the fact that both the traditional incipit of the Iliad and the alternative recorded in Apellicon’s copy fit the pattern I have just de-scribed, with, in each case, the first word an outcome of *men-/mneh2- (specifi-cally the accusative of a noun) and the second the outcome of *h2u̯eid- (specifically a present form of a verb): crudely speaking, ##Μῆνιν ἄειδε and ##Μούσας ἀείδω are the same thing, that is to say, they are made of the same Proto-Indo-European material—material that is, furthermore, wholly appropriate at the start (and end) of traditional poetry, as ample evidence in early Greek re-veals. This in itself suggests that ##Μούσας ἀείδω καὶ Ἀπόλλωνα κλυτότοξον as an incipit has a claim on antiquity (though I stress again that I do not wish to ar-gue for—or, indeed, against—its status as a specifically Iliadic opener).16

There is, however, a small synchronic linguistic difference between ##Μῆνιν ἄειδε and ##Μούσας ἀείδω whose importance has not been properly appreciated: the initial alpha of ἄειδε is short whereas that of ἀείδω is long. When we try to work out which is “correct,” the evidence points firmly to the former: for one thing, while ἀείδειν does not appear to have cognates outside Greek, if it is an Indo-European word of normal root structure, it can hardly go back to anything other than *h2u̯eid-, whose second laryngeal would regularly vocalize to ᾰ (and not to *ᾱ); and for another, instances of ᾱ̓ειδ- in Archaic poetry are in the signifi-cant minority.

15 The Odyssey ends #Μέντορι … αὐδήν## (24.548; αὐδήν and ἀοιδήν have a well-known close

association inside Greek); building on Watkins 1995:100–1, I propose in “Gods and Vowels” to read significance in the desinence of μυθησαίμην#, the final word of the proem of the Works and Days (10). Note also the phonetic link between ##Μῆνιν and enjambed #οὐλομένην at the start of the Iliad (1.1–2); see Katz 2013:182.

16 Since the traditional beginning of the Iliad is not usually regarded as hymnic (though see Nagy 2010b:119–21) and since there is no combination of *men-/mneh2- and *h2u̯eid- at the start of Homer’s other epic poem (but see n.14 above), one might, I suppose, argue that the appearance of ##Μῆνιν is accidental, i.e., that no diachronic significance is to be attributed to the fact that directly alongside ἄειδε (< *h2u̯eid-) is an opening word that reflects *men-/mneh2-. Needless to say, I do not believe that this is so.

The Hymnic Long Alpha 93

It would appear, then, that ##Μούσας ἀείδω shows metrical lengthening. That this is the usual line (to the extent that there is a usual line on a verse that has been treated as so marginal a part of the Greek corpus) is not surprising: the phenomenon of metrical lengthening in Homer is familiar enough that invoking it does not automatically require much justification.17 As it happens, I accept that the long alpha here is the result of some sort of secondary lengthening—just not one of the kinds of kunstsprachliche metrische Dehnung that are regularly pro-posed for Archaic epic. If Apellicon’s incipit were the only example of ᾱ̓ειδ-, or if the few examples that we have showed no pattern, then no further comment would be necessary. However, this is not the case.

There are seven or eight attestations of ᾱ̓ειδ- in Archaic Greek literature,18 six or seven of them in Homer(ica) and one in Theognis.19 These fall into two basic categories:

I In the first verse of a hymnic proem, immediately after an accusative of a divine name in the first foot of the hexameter, i.e., ##[–⏕]DIVINITY.ACC ᾱ̓ειδ- (5/6 examples)

a in the first-person singular present form ᾱ̓είδω (5 examples)

##Μούσας ἀείδω καὶ Ἀπόλλωνα κλυτότοξον (Apellicon’s Iliad 1.1) Of the Muses I sing, and Apollo of the famed bow.

##Ἥρην ἀείδω χρυσόθρονον ἣν τέκε Ῥείη (Homeric Hymn 12.1) Of Hera I sing, golden-throned, whom Rhea bore.

##Ἑρμῆν ἀείδω Κυλλήνιον Ἀργειφόντην (Homeric Hymn 18.1) Of Hermes I sing, the Cyllenian slayer of Argus.20

17 It is better to speak of phenomena (plural) since the reason for the initial long alpha of

(#)ᾱ̓θάνατος (and other forms in the paradigm) is different from the reason for the same in #Ἆρες (Iliad 5.31 = 455), for example. Chantraine (1973:94–112) offers a very useful classifi-cation of types of lengthening (and shortening); the standard work on metrical lengthening in Homer is Wyatt 1969.

18 A ninth example would be “ἀείδω Πριάμοιο τύχαϛ πόλεμόν τ’ ἀρίσημον,” which West (2013: 227) proposes as the opening of the Iliou persis, on the basis of Horace, Ars poetica 136–7 nec sic incipies ut scriptor cyclicus olim: / ‘Fortunam Priami cantabo et nobile bellum.’; West points out in n.3 a “similar reconstruction already in Salmasius 601B, ἀείσω Πριάμοιο τύχαν πόλεμόν τε κλεεννόν. (He thought it was from Lesches’ Little Iliad and should be in Lesbian.)”

19 In Hellenistic poetry, however, ᾱ̓ειδ- is not uncommon (Gow 1952:142 conveniently lists ex-amples from Theocritus, Callimachus, and Apollonius Rhodius), presumably an extension of its appearance in Homer(ica).

Joshua T. Katz 94

##Ἄρτεμιν ἀείδω χρυσηλάκατον κελαδεινήν (Homeric Hymn 27.1) Of Artemis I sing, golden-shafted (and) noisy.

##Ἴλιον ἀείδω καὶ Δαρδανίην εὔπωλον (Ilias parva fr.1.121 = fr. 28.1 Bernabé)

Of Troy22 I sing, and well-foaled Dardania.

[b in a form other than ᾱ̓είδω (1 contested example)

##Μήνην ἀείδειν τανυσίπτερον ἔσπετε, Μοῦσαι (Homeric Hymn 32.1) Of the long-winged Moon sing next, o Muses.]23

II Somewhere else (2 examples)

#ἀείσω (Theognis 4) “I shall sing” #ἀείδῃ (Odyssey 17.519) “[a singer who] sings”

As for the lone form from one of Homer’s two great epics, #ᾱ̓είδῃ at the begin-ning of Odyssey 17.519 (in a simile spoken by Eumaeus to Penelope),24 Chan-traine (1973:103) cites it without comment in his section on what ancient metricians called acephalous verses (i.e., ones that would appear to begin with an initial short syllable), while Wyatt (1969:182) writes that

20 An Attic lekythos from ca. 470 B.C. depicts a seated boy with a papyrus roll on which is writ-

ten ΗΕΡΜΕ | ΑΕΙΔΟ, i.e., Ἑρμῆ(ν) ἀείδω—quite possibly the beginning of Homeric Hymn 18 (see in the first place Beazley 1948:336–7, with plate XXXIV, and 1950:318–9).

21 See now West 2013:173. 22 Troy is of course a toponym rather than a theonym; for what it may be worth, it is regularly

described in Homer as ‘holy’ (cf., e.g., the formula Ἴλιος ἱρή# [Iliad 4.46+]; compare Nagy 2010b:120, with n.14, who provides on pp.119–20 his own taxonomy of the relationships among the five incipits I have classified as Ia).

23 There are brackets around Ib because the primacy of ἀείδειν is not assured in what is anyway universally regarded as a very late hymn (see, e.g., West 2012:240): it is true that the colloca-tion ἀείδειν … ἔσπετε is startling, and perhaps West (2003:216) is right to print Bothe’s εὐειδῆ or Càssola (1975:448) to conjecture ἀΐδιον; see Càssola 1975:588. The latest scholar to accept ἀείδειν here, albeit only in passing, is Olson (2012:298).

24 It may be noted that ᾰ̓είδῃ#—the same form, but with an initial short alpha—is the final word of the very next verse, Odyssey 17.520. According to Wyatt (1969:182), “[t]he reason [ᾱ̓είδ-] is hapax in Homer is that Homer already had many formulas involving ἀείδω with a short vowel, and it was only when he created a new expression that he had occasion to use the new quantity”; he refers to Hoekstra 1965:121, in which #ᾱ̓είδῃ is considered together with the im-perfect #ἤειδεν at Odyssey 8.514—both forms “detached from [their] formulaic context.”

The Hymnic Long Alpha 95

it should be clear that the explanation of ᾱ̓είδω [by which Wyatt means the para-digm, not the first-person singular form in particular] is to be found in the oppo-sition of αἰεί (ᾱ̓εί) with its development ᾰ̓εί. As ᾰ̓εί is to αἰεί (ᾱ̓εί) so ᾰ̓είδω is to x; x = ᾱ̓είδω. And once the form had been created, it was picked up for later (im-pressive) invocations.

Wyatt’s explanation could be correct (compare Pavese 1991:160 n.7). But there is no reason to believe that a nonce form in the Odyssey—one that further-more has a lengthened alpha at the start of a verse (cf., e.g., #ᾱ̓θάνατοι, etc.)25 —must be older than its hymnic counterparts, particularly when all of these aside from the one in Theognis are instantiations of an evident formula: ##[–⏕]DIVINITY.ACC ᾱ̓ειδ- (especially—perhaps exclusively—in the first-person singular form ᾱ̓είδω),26 which really is pragmatically very “impressive” by virtue of coming at the start of a poem27 (even Theognis’ likewise first-person singular form #ᾱ̓είσω, the sole example in a pentameter, is part of the invocation and marked further by being enjambed).28 It is not that the first words of a poem can-not be playful—##Μήνην ἀείδειν (whatever its antiquity may be; see n.23 above) is a delightful allusion to ##Μῆνιν ἄειδε29—but it does not seem likely that an initial formula in the inherited genre of hymning a divinity would be a locus of metrical innovation. Far more probable, then, is that ##[–⏕]DIVINITY.ACC ᾱ̓ειδ- is

25 Even Wyatt (1969:182), referring to LfrgE s.v. ἀείδω, ἀοιδιάω (Rosemarie Philipp), states of

the form at Odyssey 17.519 that “we can at best speak of the influence of the first arsis”; com-pare n.17 above.

26 This is followed by an agreeing adjective in the accusative or by καί and a noun-adjective combination, likewise in the accusative.

27 I.e., it occupies a “privileged position”—to take over a term that Rabinowitz (1987:58–65) uses to good effect in discussing a very different genre, 19th- and 20th-century European and American prose.

28 Compare Pavese 1991:160, with n.7; Wyatt (1969:182), who misses Apellicon’s incipit and does not mention Theognis, writes that “the initial ἀ- is always short, save in epic invocations” and Odyssey 17.519. (The form #ᾄσομαι, which shows up enjambed at Homeric Hymn 6.2, right after a poem-initial accusative noun phrase about Aphrodite, is—like the same form at Homeric Hymn 32.19—presumably a contraction of ᾰ̓είσομαι, found as such at the start of a number of Homeric Hymns [10.1, 15.1, 23.1, and 30.1]; cf. also #ᾄδομεν at Homeric Hymn 1.18.)

29 The ludic tradition will be continued by the first words spoken in Vergil’s Aeneid, Juno’s mene incepto (1.37), as noted by Levitan (1993); see also now Trinacty 2012.

Joshua T. Katz 96

an early as well as impressive invocation, and much more interesting than #ᾱ̓είδῃ in the Odyssey.30

But if this is so, then the question arises, Can metrical or some other form of secondary lengthening really be so old a phenomenon as to be inherited in a for-mula in our earliest Greek poetry? Yes, it can. The enterprise of linguistic recon-struction—and, in consequence, the reconstruction of poetic as well as normal language (see above all Watkins 1995)—relies on the uniformitarian principle, which has recently been formulated as follows: “Unless we can demonstrate a relevant alteration in the conditions of language use or language acquisition be-tween some time in the past and the present, we must assume that the same types, range, and distribution of language structures existed and the same types of lin-guistic change processes operated at that past time as in the present” (Ringe and Eska 2013:3 and 30). Now, sounds—both vowels and consonants—indisputably can be extended for a variety of purposes, including expressiveness, the linguistic function31 to which Calvert Watkins devoted his final paper, published in this volume, which ends with the statement that the “expressive dimension has a his-tory. It is the duty of historical linguistics to be concerned with … and to give an account of that history” (2013:251; italics in original).

One function of lengthening, for reasons that do not in the first place have to do with meter, is to express awe at something awful: at a snake, for example, as in one of Watkins’s favorite words, ὄπφιϛ ‘SNAKE’ in Homer and Hipponax, and (in my opinion) in the geminate -ll- of the first element (‘eel’) of the mon-strous Hittite compound illuyankaš, the Proto-Indo-European preform of whose second element (‘(water) snake’) almost certainly contains the extraordinarily marked combination of the rare vowel *a and the rare consonant *gu̯h.32 But an- 30 West (1997:170–3) suggests that “[w]hen the early Greek poet … begins by announcing his

theme … [by] say[ing] ‘I will sing’ or ‘Let me sing’ of N” (170), this is a borrowing from the (non-Indo-European) Near East, e.g., from the use of azammur and luzmur in Akkadian. This is not impossible, but I am especially skeptical as far as the foundational formula ##[–⏕]DIVINITY.ACC ᾱ̓ειδ- is concerned. I note that two of the many passages that West cites in the footnote to the phrase just quoted (1997:170 n.5), Homeric Hymn (2) to Demeter 1 and PMG 885.1 (both with ᾰ̓ειδ-), reappear in West 2007:305–6, where the matter at hand is spe-cifically Indo-European ways of calling to a god.

31 Compare Jakobson 1960:354 on the “emotive” function. 32 In this volume, Watkins mentions both αἰόλον ὄπφιν at Iliad 12.208 (see also, e.g., Watkins

1995:29, and compare 461–2) and the expressive power of *a and *gu̯h (compare, e.g., Wat-kins 1995:305, with n.1); for illuyankaš, see Katz 1998. In Katz 2006:178–9, with n.55 (where further literature on such matters in Greek is cited), I propose that the iota in the stem of the

The Hymnic Long Alpha 97

other function is to express awe at something awesome, and this is what I believe we see—or rather, in oral poetry, experience aurally—in the long alpha of ᾱ̓ειδ- in ##Μούσας ἀείδω and its ilk.

In “Gods and Vowels,” I make the suggestion that the Sanskrit syllable om (< aum), which is both sacred and attention-getting, finds a Greek analogue, func-tionally and phonetically, in ##Μῆνιν ἄειδε and μνήσομ’ ἀοιδῆς##—and hence also in the apparently inherited juxtaposition of the roots *men-/mneh2- and *h2u̯eid-. Note that om, like these Greek collocations, opens and closes mantras and also contains in very little space both a range of vowels (a series of sounds I connect to the divine) and a nasal hum. My suggestion may be judged overbold by some, and so may my further idea that om, which is the only word in Sanskrit that is regularly pronounced overlong, with so-called pluti, has a correspondent in the functionally and phonetically marked scansion of the Greek reflex of *h2u̯eid- as ᾱ̓ειδ- at the start of a hymn, as in ##Μούσας ἀείδω. Be that as it may, my con-tention in the present paper is that ᾱ̓ειδ- in hymnic incipits, though evidently re-quired by the meter, is no standard (low-level and minimally interesting) example of metrical lengthening. Rather, the etymologically unexpected additional mora in the formula ##[–⏕]DIVINITY.ACC ᾱ̓ειδ-, especially with the first-person singular form ᾱ̓είδω, is an inherited way—how deeply inherited it is impossible to know in the absence of known cognates of ἀειδ- outside Greek—for the bard to bring his audience to attention at the start of an expression of divine praise.33

Greek designation of the Sphinx, (Σ)Φῑκ-, shows expressive lengthening. Schnoebelen 2012, a recent dissertation on linguistics and emotion that considers such matters as the addition of “unnecessary” letters in tweets (e.g., yayyyy), has received national attention (see, e.g., Doll 2013) and should incite greater interest in expressivity among linguists, synchronic and dia-chronic and phonologically and pragmatically oriented alike.

33 Wyatt (1969:243) may place too much emphasis on book divisions when he writes of the ha-pax λῦτο at Iliad 24.1 that “[w]hen Homer wished to signal the beginning of the impressive conclusion to his Iliad, he did so by means of a metrically lengthened form: λῦτο δ’ ἀγών”; as for his claim that “one would not expect to find an archaism in the opening line of a highly dramatic and thematically important book” (209), it should be clear that I reject in principle the idea that lengthening in an incipit cannot be archaic. That said, I am not unsympathetic to the tenor of Wyatt’s overall conclusions about metrical lengthening:

Metrical lengthening was an integral part of the texture of oral epic poetry. Homer’s listeners expected to hear him sing the distant and magnificent deeds of their own ancestors in a different, more exalted, though still comprehensible, form of their own language. And metrical lengthening was one of the familiar devices which served to create this aura of remoteness. For this reason it is best to think of these artificially

Joshua T. Katz 98

As I point out in “Gods and Vowels,” the run of the vowels from (long) alpha to (long) omega in ᾱ̓είδω is remarkable in itself as well as being a sign of the di-vine—not too far before the 8th century, the word would have been pronounced ᾱ̓ϝείδω and have thus presented all five qualities of (semi)vowel in just six sounds: ĀVEI-Ō—with the opening sound, [aː], the most resonant and (I would say) therefore the most godly vowel in the language,34 and an iconic one here, to boot, since it literally sings in a word that means ‘I sing’. This hymnic long alpha, as I call it, is in effect an ideophone, representing the idea of the sacred in its very sound.35 Acting as a call to prayer, it neatly brings together salient features of “the Hwæt paradigm” (Foley 1991:214–23) with which Beowulf and a substantial number of other Old English poems begin,36 the drawn-out English word gawd,37 and the melodious sacred amen##38—only uttered at the start of liturgy rather than the end.39

lengthened vowels not as cases of metrical lengthening, a term tied too much to me-chanical processes of creation, but rather as examples of the epic lengthened grade, an expression which both relates the phenomenon to its linguistic causes, and also removes it from the domain of sheer necessity into the realms of art. (243)

34 Compare Stanford 1967:24–5 n.51, 51–2, and 82. 35 The classic and oft-repeated definition of “ideophone” comes from Doke 1935:118: “A vivid

representation of an idea in sound. A word, often onomatopoeic, which describes a predicate, qualificative or adverb in respect to manner, colour, sound, smell, action, state or intensity.” Ideophones, which frequently involve reduplication or extra-long vowels, have historically been associated with African languages, but there has recently been a good deal of cross-linguistic research: see, e.g., the papers in Voeltz and Kilian-Hatz 2001a, including the intro-duction of the editors, who note that “[i]n almost all cases investigated, … ideophones have a particular often special phonology … [and] are very often only used in oral language” (Voeltz and Kilian-Hatz 2001b:2). See also Mark Dingemanse’s blog: ideophone.org.

36 “[T]he Hwæt paradigm for signaling the opening of a heroic narrative consists regularly of the interjection, a verb of speaking or hearing, and identification of the speaker as ‘we’ or ‘I’” (Fo-ley 1991:222).

37 The form gawd earns a self-standing entry in the OED (“Vulgar or slang pronunciation”), where it is noted s.v. god, n. that “[f]rom a desire to utter the name of God more deliberately than the short vowel naturally allows, the pronunciation is often /gɒːd/ or even /gɔːd/.”

38 According to the Talmud (Berakhot 47a), drawing out amen prolongs life (and saying it quick-ly shortens it). Note also the practice of the Marcosians (a Gnostic sect active in France from the 2nd to 4th centuries A.D.) of chanting in unison amen and St. Jerome’s rhetorical question about the Romans in the preface to Book 2 of his Commentary on Galatians: ubi sic ad simili-tudinem cœlestis tonitrui Amen reboat, et vacua idolorum templa quatiuntur? (PL 26, col. 381) “Where else [but in Rome] does ‘Amen’ reverberate like thunderclaps in the sky, and

The Hymnic Long Alpha 99

References

Beazley, J. D. 1948. Hymn to Hermes. American Journal of Archaeology 52.336–40 + plates XXXIV–XXXVIII.

———. 1950. Some Inscriptions on Vases: V. American Journal of Archaeology 54.310–22.

Cain, Andrew. 2010. St. Jerome, Commentary on Galatians. Washington, DC: Catholic University of America Press.

Càssola, Filippo. 1975. Inni omerici. Milan: Mondadori. Chantraine, Pierre. 1973. Grammaire homérique I: Phonétique et morphologie5. Paris:

Klincksieck. Doke, Clement M. 1935. Bantu Linguistic Terminology. London: Longmans, Green. Doll, Jen. 2013. Why Twitter Makes Us Want to Add Extra Letterssss. The Atlantic Wire

(21 February). http://www.theatlanticwire.com/entertainment/2013/02/why-twitter-makes-us-want-add-extra-letterssss/62348/.

Foley, John Miles. 1991. Immanent Art: From Structure to Meaning in Traditional Oral Epic. Bloomington: Indiana University Press.

Ford, Andrew. 1992. Homer: The Poetry of the Past. Ithaca: Cornell University Press. Gow, A. S. F. 1952. Theocritus2 II: Commentary, Appendix, Indexes, and Plates. Cam-

bridge: Cambridge University Press. Hoekstra, A. 1965. Homeric Modifications of Formulaic Prototypes: Studies in the De-

velopment of Greek Epic Diction. Amsterdam: Noord-Hollandsche Uitgevers Maatschappij.

Jakobson, Roman. 1960. Closing Statement: Linguistics and Poetics. In Thomas A. Sebeok (ed.), Style in Language, 350–77. Cambridge, MA: Technology Press of Massachusetts Institute of Technology. [Reprinted in Selected Writings III: Poetry of Grammar and Grammar of Poetry, ed. by Stephen Rudy, 18–51. The Hague: Mou-ton, 1981.]

Katz, Joshua T. 1998. How to Be a Dragon in Indo-European: Hittite illuyankaš and Its Linguistic and Cultural Congeners in Latin, Greek, and Germanic. In Jay Jasanoff, H. Craig Melchert, and Lisi Oliver (eds.), Mír Curad: Studies in Honor of Calvert Wat-kins, 317–34. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.

———. 2005. The Indo-European Context. In John Miles Foley (ed.), A Companion to Ancient Epic, 20–30. Malden, MA: Blackwell.

where else are the empty shrines of false gods shaken to the core?” (translation: Cain 2010:131–2). When chanted or sung, the word is of course often extended, and in conservative liturgical tradition in English, the first vowel is pronounced [ɑːmɛn] rather than [eɪmɛn], as though the Great Vowel Shift never happened.

39 As for the long alpha (rather than eta) of θεᾱ́ in the incipit of the standard Iliad, that is a sub-ject for another occasion.

Joshua T. Katz 100

———. 2006. The Riddle of the sp(h)ij-: The Greek Sphinx and Her Indic and Indo-European Background. In Georges-Jean Pinault and Daniel Petit (eds.), La Langue poétique indo-européenne: Actes du Colloque de travail de la Société des Études In-do-Européennes (Indogermanische Gesellschaft/Society for Indo-European Studies), Paris, 22–24 octobre 2003, 157–94. Louvain: Peeters.

———. 2010. Inherited Poetics. In Egbert J. Bakker (ed.), A Companion to the Ancient Greek Language, 357–69. Malden, MA: Wiley-Blackwell.

———. 2013. Saussure’s anaphonie : Sounds Asunder. In Shane Butler and Alex Purves (eds.), Synaesthesia and the Ancient Senses, 167–84. Durham: Acumen.

———. Forthcoming. Gods and Vowels. To appear in José Virgilio García and Ángel Ruiz (eds.), Poetic Language and Religion in Greece and Rome. Newcastle: Cam-bridge Scholars Publishing.

Levitan, William. 1993. Give up the Beginning?: Juno’s Mindful Wrath (Aeneid 1.37). Liverpool Classical Monthly 18.14.

Nagy, Gregory. 1998. The Library of Pergamon as a Classical Model. In Helmut Koester (ed.), Pergamon: Citadel of the Gods. Archaeological Record, Literary Description, and Religious Development, 185–232. Harrisburg: Trinity Press International.

———. 2004. Homer’s Text and Language. Urbana: University of Illinois Press. ———. 2010a. Homer Multitext Project. In Jerome McGann (ed.), Online Humanities

Scholarship: The Shape of Things to Come. Proceedings of the Mellon Foundation Online Humanities Conference at the University of Virginia, March 26–28, 2010, 87–112. Houston: Rice University Press. Also available online: http://cnx.org/ content/m34314/1.2/ (last updated 14 May 2010) and http://chs.harvard.edu/wa/ pageR?tn=ArticleWrapper&bdc=12&mn=4087.

———. 2010b. Homer the Preclassic. Berkeley and Los Angeles: University of Califor-nia Press.

Olson, S. Douglas. 2012. The “Homeric Hymn to Aphrodite” and Related Texts: Text, Translation and Commentary. Berlin: de Gruyter.

Pavese, C. O. 1991 [1993]. L’inno rapsodico: Analisi tematica degli inni omerici. A.I.O.N., Sezione filologico-letteraria 13 [= Albio Cesare Cassio and Giovanni Cerri (eds.), L’inno tra rituale e letteratura nel mondo antico: Atti di un colloquio, Napoli, 21–24 ottobre 1991, Rome: Gruppo Editoriale Internazionale].155–78.

Pelliccia, Hayden N. 1997. As Many Homers as You Please. New York Review of Books 44/18 (20 November).44–8.

Rabinowitz, Peter J. 1987. Before Reading: Narrative Conventions and the Politics of Interpretation. Ithaca: Cornell University Press.

Ringe, Don, and Joseph F. Eska. 2013. Historical Linguistics: Toward a Twenty-first Century Reintegration. Cambridge: Cambridge University Press.

Schnoebelen, Tyler Joseph. 2012. Emotions are Relational: Positioning and the Use of Affective Linguistic Resources. Ph.D. diss., Stanford University.

The Hymnic Long Alpha 101

Stanford, W. B. 1967. The Sound of Greek: Studies in the Greek Theory and Practice of Euphony. Berkeley and Los Angeles: University of California Press.

Trinacty, Christopher. 2012. The Manipulation of Juno’s μῆνις: A Note on Lucan’s BC 9.505 and Silius Italicus’ Pun. 12.284. Illinois Classical Studies 37.167–73.

Voeltz, F. K. Erhard, and Christa Kilian-Hatz (eds.). 2001a. Ideophones. Amsterdam: Benjamins.

———. 2001b. Introduction. In Voeltz and Kilian-Hatz 2001a, 1–8. Watkins, Calvert. 1995. How to Kill a Dragon: Aspects of Indo-European Poetics. New

York: Oxford University Press. ———. 2013. Aspects of the “Expressive Dimension” in Indo-European: Toward a

Comparative Grammar of Speech Registers. In this volume, 243–53. West, M. L. 1966. Hesiod, “Theogony.” Oxford: Clarendon Press. ———. 1978. Hesiod, “Works & Days.” Oxford: Clarendon Press. ———. 1997. The East Face of Helicon: West Asiatic Elements in Greek Poetry and

Myth. Oxford: Clarendon Press. ———. 1998. Homeri “Ilias” I: Rhapsodias I–XII continens. Stuttgart: Teubner. ———. 2001. Studies in the Text and Transmission of the “Iliad.” Munich: Saur. ———. 2003. Homeric Hymns; Homeric Apocrypha; Lives of Homer. Cambridge, MA:

Harvard University Press. ———. 2007. Indo-European Poetry and Myth. Oxford: Oxford University Press. ———. 2011. The Making of the “Iliad”: Disquisition and Analytical Commentary. Ox-

ford: Oxford University Press. ———. 2012. Towards a Chronology of Early Greek Epic. In Øivind Andersen and Dag

T. T. Haug (eds.), Relative Chronology in Early Greek Epic Poetry, 224–41. Cam-bridge: Cambridge University Press.

———. 2013. The Epic Cycle: A Commentary on the Lost Troy Epics. Oxford: Oxford University Press.

Wyatt, William F., Jr. 1969. Metrical Lengthening in Homer. Rome: Edizioni dell’Ate-neo.

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ]>> setdistillerparams> setpagedevice