CC BY
39
А. С. Николаев
A NEW ARGUMENT AGAINST THE ASSUMPTION OF AN «AEOLIC PHASE»
IN THE DEVELOPMENT OF HOMERIC EPIC DICTION*
As is well known, the language of Homeric epics contains a veritable host of Aeolic words and morphemes, many of which have a different metrical structure from their Ionic counterparts (or do not have equivalents in Ionic) and are deeply embedded in the formulaic system of Homeric dialect. While there is little disagreement that Homeric forms such as поті, K£(v), по&єот, пюирє^, пє^пшроЛа or derive
from an Aeolic source, there are two schools of thought as to how exactly these Aeolic elements penetrated the Homeric epic diction. The more traditional view is that the Ionian epic tradition as we find it in Homer was preceded by an Aeolic phase and traditional stories, inherited from early Mycenaean times, were transmitted to the Ionian singers by Aeolians.1 Under this theory the «Aeolisms» of the Homeric language are, in M. L. West's words, «a residue left after Ionian poets had adapted an Aeolic poetic language into their own dialect as far as it would go» (West 1988: 162). More recently, however, an alternative view has been gaining ground,
* Symbols: |,5 penthemimeral caesura, |" tritotrochaic caesura, |,7 hepthemimeral caesura, I I bucolic dieresis, - = long, ^ = short, ^ = anceps, = bridge. A number (1-6) above a vowel in the princeps indicates the foot. In using "princeps" and "biceps" instead of "arsis" and "thesis" I follow West 1982 (cf. Зайцев 1999: 5). I am deeply grateful to M. Peters without whose teaching and example of philological rigor this note would not have been written.
1 Ever since Ritschl 1838: 59-60 an Aeolic phase has been postulated
e.g. by Parry 1932; Wathelet 1970; Janko 1982; West 1988; Forssman
1991: 273; Ruijgh 1995; Berg-Haug 2000. See the recent surveys by
Казанский 2004 and Hackstein 2010.
namely, that the Ionic epic tradition existed side by side with an Aeolic epic tradition and the «Aeolisms» are a result of borrowing and diffusion, rather than remnants of a preceding phase. Whenever the metrical structure of original Ionic forms was distorted by the sound changes or Ionic simply did not have a form needed to fill a particular metrical slot, the bards were able to borrow Aeolic forms from the parallel poetic tradition.2 The purpose of this paper is to discuss one hitherto unnoticed argument in favor of the second view.
One linguistic feature that has figured prominently in the debate is the form of thematic active infinitives in Homer. In proto-Greek thematic present stems and thematic («second») aorists had the infinitive suffix *-ehen and this state of affairs is continued in Attic-Ionic, West Greek and «Achaean» dialects (cf. Mycenaean e-ke-e /(h)ekhehen/).3 By the loss of intervocalic *-h- and vowel contraction *-ehen gave Attic-Ionic -£iv (i.e. /-en/, with a «spurious diphthong»), which is also the most frequently attested thematic infinitive suffix in Homeric dialect.4
However, in Homer we also find thematic infinitives in -£^£v (e.g. inf. pres. eAke^ev, inf. aor. £A0£^£v). These forms have traditionally been viewed as Aeolic elements of the Homeric language, no doubt, correctly, since among all dialects of Ancient Greek, only in mainland Aeolic (or perhaps already in Proto-Aeolic5) was the originally athematic
2 For the thesis of concurrent existence of Aeolic and Ionic epic traditions see e.g. Drerup 1903: 107; Hooker 1977: 70-82; Miller 1982: 34 et passim; Peters 1989: 6 et passim; Horrocks 1987; 1997: 214-17; Wyatt 1992; Hajnal 2003: 81.
3 On the dialectal outcomes of *-ehen see Garcia-Ramon 1977.
4 See e.g. Meyer 1856: 5-7. It should be noted that there are formulae in Homeric epics where the ending of the thematic infinitive is used in princeps: in these cases -eiv is of course irresolvable and it is impossible to restore an «original» Aeolic ending -e^ev. The ending -eiv thus belongs to the inherited Ionic lore. This observation may by itself constitute an argument against the assumption of a special «Aeolic» phase in the development of epic diction (cf. Peters 1989: 6).
5 West Thessalian and Lesbian have thematic infinitives of the type ^epev / ^ep^v, but these forms do not have to reflect the original Proto-Aeolic situation: see Garcia-Ramon 1975: 90.
infinitive suffix -^£v6 transferred to thematic stems: from epichoric Aeolic one may compare East Thessalian unapx£-^£v, ^£v£^£v, Boeotian KaxapaAA£^£v, kqi&&£^£v7 and an infinitive ^]£Q£^£v is found in Corinna (fr. 654.20 PMG).8
The use of thematic infinitives in -£^£v in Homeric epics shows a remarkable tendency: the suffix of these forms for the most part occupies the biceps of the fourth or the fifth foot (before a vowel): see the material presented in Chantraine 1973 [1958]: 490-2. This distribution of forms in -£^£v conforms to the well-known preference for dactylic word-end before the bucolic dieresis and before the sixth foot.9 The question is then, why these Aeolic forms appear precisely under such metrical conditions. There are two possible explanations.
1) Forms such as inf. pres. eAk£^£v or inf. aor. £A0£^£v should be viewed as remnants from the «Aeolic stage» of epic diction: these forms were preserved in old formulae and they were not «ionicized» by Ionic bards, because Ionic -en in IAk£iv or £A0£lv would then fill the biceps, resulting in an undesirable spondaic fourth (or fifth) foot. In other metrical contexts (e.g. the first or second foot) the pressure for a dactyl
6 *-men was probably originally used only with monosyllabic athematic stems: see Peters 1986: 308.
7 See Bechtel 1921: 193, 289; Thumb-Scherer 1959: 43, 71; Wathelet 1970: 319; Blumel 1982: 208-10. I am not sure I understand Haug (2002: 157), who objects to Horrocks 1987 on the grounds that inf. -e^ev is also found in Doric dialects and the extension of the suffix -^ev to thematic stems «est assez ancienne en grec»: at least, I am not aware of any such forms in any Doric dialects, save for Cretan npopeine^ev on which see the next footnote.
8 We do not have a single inscriptional example of thematic inf. aor. (except i&e^ev SEG 37.389 found on a Boeotian lead tablet dating from the 2nd cent. AD), but it is very likely that athematic -^ev, used with presents and aorists alike, was transferred to both thematic presents and aorists in Aeolic. Indirect support for this claim may come from Cretan njpopeine^ev (IC I xviii 1), if the ending -^ev in this form should be attributed to an Aeolic substrate (however, this is not the only possible explanation of the form, see Garcia-Ramon 1973: 268-9).
9 That is, the avoidance of bucolic bridge: -,4 -^,5 , and Meister's bridge: -,5 - -,6, resp. See Bekker 1863: 144-8; Witte 1915: 484; Meister 1921: 7-9, 12-22; Witte 1972: 88-90, Nagy 1974: 74-5.
was felt less acutely and the singers were able to replace traditional infinitives in -£^£v with their native forms ending in -£iv.
2) Forms such as inf. pres. eAk£^£v and inf. aor. £A0£^£v are products of secondary «aeolicization» of the Homeric epic diction: in cases where *-ehen in traditional (Ionic) formulae occupied the biceps of a dactylic fourth or fifth foot, the singers had an option of substituting Aeolic -£^£v for the metrically inept Ionic -en. In the first, second, third or last foot of Homeric hexameter a spondee was more admissible and hence original Ionic forms in -£iv could be retained there.
In order to decide between these two scenarios we must take a closer look at the distribution of the infinitives in -£^£v. While the majority of these forms fill a dactylic fourth or fifth foot (extending into the biceps of the preceding foot in case of longer forms), infinitives with the suffix -£^£v are also made from thematic stems derived from the roots of the structure (C)VC: ay-, £%-, [^£A-, ^£v-, ^£Q-. Even though the infinitives ay£^£v, ^£Q£^£v, etc. had a metrical structure different from eAk£^£v or tc£^tc£^£v, such forms could of course still be used in hexameter: for instance, they may occupy the metrical slot after the penthemimeral caesura, cf. K 411:
5iX0a&ia^ K^pa^ |, ^£p£^£v 0avaxoio t£Ao^ &£
In fact present infinitives of this type are reasonably well attested in Homer.10
The situation is vastly different, however, when we consider thematic aorists made from roots of the same type: aorists of the type ^ay£^£v are exceptionally rare and are for the most part found in the Odyssey.11 Instead we find a robust
10 dye^ev A 323, 443, H 418, 420, 471, T 195, £ 274; exe^ev A 302, E 492, I 708, N 2, P 476, T 148, 5 419, n 313, 0 529; ea0iDe^ev |3 305, $ 69; fieve^ev E 486, O 400, i 97, u 330; ^epe^ev E 234, I 411, Q 119 ( = 147, 176, 196); fieAe^ev a 421.
11 They are: ^aye^ev k 386, o 378, n 143, p 404, a3; xpa^e^ev H 199, E 436, y 28 (always in VE formula Yevea0ai xe xpa^e^ev xe); ne$pa5e^ev n 49; me^ev o 378, n 143, a 3 is found in the same verse with $aY£^ev (Augenblicksbildung?), but the form is also found at n 825.
group of aorist infinitives ending in -££iv: paA££iv, ^ay££iv, 0av££iv, L&££iv, 6aK££iv, eA££lv etc., with 102 attestations from 19 verbal roots on the whole. The majority of these forms are found before consonant; whenever they stand before a vowel, they are almost invariably followed by a caesura. They usually are located in the biceps of the 3rd foot and the princeps of the 4th foot or in the biceps of the 2nd foot and the princeps of the 3rd foot (in rare cases -£iv fills the princeps of the 5th foot). In other words, we find aorist infinitives in this mysterious -££iv precisely where the proponents of the «Aeolic phase» would expect to find infinitives in -£^£v.
It behooves us to discuss the origin of the ending -££iv. According to the usual view (that goes back to G. Curtius apud Renner 1868: 34) these forms are a product of Attic ^£TaxapaKTnpLCT^o^, viz. a transposition of OYrEEN written in an Old Ionic alphabet (with a metrical structure ^ ^ - before a consonant) into ^uy££iv with the ending of Attic ^uy£iv.12 But if this indeed were the process responsible for the creation of Homeric ni££iv, etc., one would expect that present stems of the type ^£pw (inf. *pherehen, spelled in Old Ionic alphabet as OEPEEN) would show such distraction, too, which is precisely not the case: simple thematic present stems never take this ending. For the same reason it is hard to believe the theory advanced by Mangold 1873: 204 («-££iv errori rhapsodorum recentiorum tribuendum, cui jam tribuimus formas in -owvt-et -o^-»): while there is no problem in assuming a «diachronically false» distraction, Mangold's theory fails to explain the absence of present infinitives of the type *^£Q££iv.
The correct explanation was given by Chantraine (1973 [1958]: 493) who (somewhat implicitly) suggested viewing the suffix -££iv as a product of a four-part analogy to contracted and uncontracted infinitive forms from present stems in -ee/o-(type ^OQ£Lv : ^OQ££iv)13:
12 Rzach 1875-76: 441; Simmerle 1874: 8; Hermann 1914: 262; Schwyzer 1939: 807; Wuest 1943: 99.
13 Smyth 1894: 499 generally subscribes to Renner's view (see above), but adds a wise (if confusing) provision «false transcription of EEN through the analogy of $iAeeiv and congeners».
present stems in -ee/o-14
-£IV
-££IV
thematic aorists in -e/o-
-£LV
-X
where X is resolved as -££iv
It now becomes clear why the presents of the type ^£pw (inf. ^£Q£iv) never have such distracted infinitives: they never had a circumflex suffix -£iv.
It is not quite certain whether we should see the suffix -££iv as an analogical creation within the Homeric Kunstsprache or as a real form, current in Ionic dialect(s). This suffix is not found in epichoric Ionic, it is exceedingly rare in elegiac poetry and in prose it demonstrably has a pseudo-Ionizing function.15 Nevertheless, it cannot be excluded that we are dealing with a spoken language form that went into oblivion in later Ionic, both literary and colloquial. One thing is certain: it is a specifically Ionic formation16 which cannot be of Aeolic provenance.17
Now we can return to the peculiar distribution of thematic infinitives in Homeric epics and the two scenarios suggested above: «Aeolic phase» vs. «aeolicization».
Under a theory of an earlier «Aeolic phase» of Homeric epics it is simply impossible to adequately capture the distribution of the thematic aorist forms: why would a bard replace traditional Aeolic -£^£v (ex hypothesi the only active infinitive ending in all thematic forms) by -££iv specifically in
14 Including simple thematic verbs of the structure *CRe(h/u)e/o-, e.g. Хеш, inf. xeeiv ~ X£lv.
15 All -eeiv aorist infinitives in Herodotus are Homeric, see Smyth 1894: 499.
16 The grammarians often call the infinitive ending -eeiv specifically Ionic (e.g. Et.M. 465.49-50: <’I5eeiv>: [...] ol Twveg era. xwv eig eiv dnape^dxwv, nAeovdCovxeg xo e, bid xou eeiv профероист1, 0avelv, Gaveeiv), but it is very likely that « T^veg » here stands for Ionic hexameter poetry.
17 In Thessalian and Boeotian contract verbs of the type фореш have thematic active infinitives in -ei^ev, while in Aeolic of Asia Minor these forms end in -qv.
the third/fourth foot before a consonant or a caesura? For
3
instance, N 629 пир oAoov paA££, iv KT£ivai 5' ^рыа^ А%аюи^ has no metrical advantage over its alleged Aeolic
3
predecessor *пир oAoov paA£^£, v KT£ivai 5' ^рыа^ A%aiou^
4
and similarly *5oid> 5' ou 5uva^ai (p)L5£^£, v кос^тор£ Aawv is not metrically inferior to the actually transmitted Г
4
236 5oid> 5' ou 5uva^ai L5££, iv кост^^тор£ Aawv. If — according to the «Aeolic phase» theory — the infinitives in -£^£v were able to survive in archaic formulae, there is no reason why they should not have likewise survived in contexts as above.18 The aorist infinitives of the type paA££iv are thus very unlikely to be replacements for an older Aeolic *paA£^£v.
It remains to test the «aeolicization» theory against the same facts. It can easily be seen that this theory accounts for Homeric present and aorist infinitives of the type eAk£^£v and eA0£^£v (- ^ w) at least just as well as the «Aeolic phase»
1 _ 2 _ 3
theory does: for instance, a formula */na, uon o, kuporo, n
4
skhedon e, lthehen/ 'go to quick-going ships' after the loss of *-h- and subsequent vowel contraction came to contain an undesirable spondee before the bucolic dieresis (*/n£uon OkuporOn skhedon elthen/ - - - ^ ^ - ^ ^ -,4 - ||), but by borrowing the Aeolic form from a parallel tradition the singers were able to restore a dactyl in the fourth foot (cf. К 308+ # v^wv ыкипбр^у ctx£5ov M.0£^£v ek t£ nu0£a0ai). According to the diffusionist approach, a very similar motivation is behind the appearance of Aeolic genitives -ao, -a«v as replacements of the archaic Ionic -^o, -^wv.19
18 Hermann (1914: 262) suggested that «im Aolischen hatte man im Infinitiv des Prasens bei thematischen wie bei athematischen Stammen -^ev, -^evai, im Aorist aber nur bei den athematischen; die thematischen bildeten -eev», but this theory lacks conviction and is not supported by the facts (see above, Fn. 8).
19 See e.g. Horrocks 1987: 290.
No doubt, there must also have been traditional formulae where the infinitive (for instance, *(h)ekhehen) conformed to an anapaestic foot before a consonant (^ ^ -); compare exempli gratia reconstructions:
*/(h)ekhehen lauon kwe (h)ikwons kwe/20 or
5
*/(h)ekhehen patrouiia panta/ (|, ^ ^--------^ ^ -^ #).
In this case, yet again, the sound change (*(h)ekhehen > Old Ionic /ekhen/) must have distorted the metrical structure of the
verse, resulting in a defective third foot (- |, -----^ ^ -^ #).
Two strategies were available: either the formula could be modified so that it could be used after the feminine caesura (compare n 388 Kai £X£iv naxpwia navia 'and hold his father's property'21) or an Aeolic variant with a suffix -£^£v could be used instead (compare I 708 KapnaAi^w npo vewv
5
|, £X£^£v Aaov T£ Kai Innou^ 'forthwith bring out your host
5
of warriors and your horsemen' or n 476 ltctcwv a0avaxwv |, £X£^£v S^oxv T£ ^£vo^ T£ 'have control and power over the immortal steeds').22
20 For correlative *kwe ... *kwe in this reconstruction of the verse compare dv5pwv xe 0ewv xe, -5^^og xe noAig xe, etc.
21 Compare also p 22 exov |, naxpwia epYa. For rcaxpwia navxa occupying the metrical slot from the penthemimeral caesura to the princeps of the fifth foot compare p 80; u 336; x 61.
22 Of course, Aeolic forms with a different prosodic structure were not available in many situations where sound change affected the original scansion. This is the case with a likely archaic formulaic expression that violates both the bucolic and Meister’s bridge in Homer, namely, qw 5Tav, where spondaic qw is nine times located in the 5th foot and twice in the 4th foot: the accusative of 'dawn' (Proto-Greek *auhohs < *h2eusosm) was disyllabic in Ionic and Aeolic alike and borrowing the form from the neighboring poetic tradition would not remedy the situation in this case. K 574 i5pw noAAov in the Doloneia may or may not be a similar case, see Meister 1921: 182.
Another example of a spondaic fifth foot presents a trickier problem. A 639 aiYeio,5v Kvn xu,6pov may well reveal an old expression where the biceps of the fifth foot was filled by a disyllabic form *knee < *knehe < *kneie (Schmidt 1968: 29-31) or *knae < *knahe < *knaie (West 1998): but whatever the exact protoform should be, both
The question to be answered now is: why (1) the Ionic reflex of inf. pres. *(h)elkehen was replaced by its Aeolic counterpart eAk£^£v, (2) the Ionic reflex of inf. aor. *elthehen was replaced by £A0£^£v and (3) the Ionic reflex of inf. pres. *ekhehen was replaced by £x£^£v, but inf. aor. *gwalehen was never replaced by *paA£^£v in Homer?
One (and I think, the only) sensible answer to this question is the following: when the need arose to find a replacement for metrically undesirable forms of the type paA£iv ( < *gwalehen) that in traditional formulae were supposed to fill a metrical slot ^ ^ -,3/4/5, the bards had no need to borrow synonymous forms from the Aeolic tradition, since they were able to use the forms of the type paA££iv, whether because such forms belonged to their native dialect or because these forms were familiar to them from the poetic tradition they belonged to. But they could not use the Ionic ending -££iv if the traditional formula contained a dactylic aorist infinitive (*elthehen > Ionic eA0£lv, *pephnehen > Ionic TC£^v£iv), since such forms would have had a cretic shape (*£A0££iv - ^ - ), and therefore the singers had to resort to Aeolic aor. inf. £A0£^£v and TC£^v£^£v. The reason why they did not have an option of substituting a form in -££iv for Old Ionic present infinitives
Ionic and Aeolic forms of this verb must have been monosyllabic by the time the Homeric text was taking its form and once again, «aeolicization» would be of little avail. (West, who cites Boeotian xupOKvaCTCTxi&eg SEG 24.361 as a piece of evidence for the root form
*kna-, plausibly interprets kv^ as an Atticism of the transmission: and yet, if *kna- is indeed the correct reconstruction, one wonders whether we might in fact be dealing with a Boeotian contraction product of *-ae- such as, according to M. Peters (p.c.), is also found in Homeric ^pi < *aieri; as far as Boeotian elements in the (West)Ionic component of Homeric diction are concerned, Peters reminds me of gen. sg. dag, Zenodotus' reading in 0 470). However, it is also possible to argue that *knee ( < *kn§e < *knae) is an artificial thematic form, based on an athematic 3 sg. *kna < *kneh2-t and created by the singers precisely in order to fill the biceps (Hackstein 2002: 119); in this case kv^ does not constitute a violation of Meister's bridge and the transmitted kv^ would in this case be purely orthographic.
(*pherehen > Ionic ^£p£iv) should now be clear: there has never been a pres. inf. of the type *^£p££iv in their repertoire.
Not only does this solution explain why aorist infinitives in -££iv are only formed from stems of a specific prosodic structure, but we can now also understand why these forms are almost never found before a vowel: in original hexameter an aorist infinitive of the type *gwalehen could not have been used before a vowel, since the form would then scan as ^ ^ The predictable absence of such metrical contexts explains why the aor. inf. in -££iv (itself a replacement of original *-ehen) is only found before a consonant or a caesura.23 The facts fit together as though dovetailed: the «aeolicization» theory is able to explain all the details of the distribution of active thematic infinitives in Homer.
The small section of Homeric grammar just surveyed has interesting implications for the proper appreciation of the other major representative of early hexameter poetry, namely Hesiod. It cannot be entirely fortuitous that Hesiod in Works and Days and Theogony uses only forms of the second aorist infinitive with contraction: a^ipaAdv E. 787, £Ltc£lv E. 453, £vion£Lv Th. 369, £^£A0£lv Th. 772, nap£A0£iv Th. 613, E. 216, L&£lv Th. 701, 0av£iv E. 175, 687, rcapaox£Lv E. 712, xa^£iv E. 807, npo^uy£iv E. 647. In all cases -£iv is either located in the princeps or fills the biceps of the 6th foot. Not a single aorist form in -££iv is attested.
23 It should be noted, however, that there was a way of using forms of the type *gwalehen in hexameter before a vowel, namely, by a metrical lengthening of the first vowel. That this was indeed the case, can perhaps be proved by n 825 rae^ev d^w = n 143 rne^ev auxwg (in both cases -,5 ^ -,6 - #): aor. rae^ev is here an aeolicized version of metrically lengthened *pi(i)ehen, similar to the case of eA0e^ev substituted for dactylic *elthehen (one should bear in mind, nevertheless, that the lengthening of *pi(i)ehen to *pi(i)ehen had support elsewhere in the paradigm of this verb, e.g. in aor. ipv. ra0i).
Things are different with Pseudo-Hesiodea, not unexpectedly. The Shield of Heracles (the Scutum) that in the antiquity was attributed to Hesiod, but is more likely to have been composed in the beginning of the 6th century,24 shows the following forms: £A££iv (337), Ain££iv (332; before a
consonant), ni££iv (252), npa0££iv (240; before a consonant) and ^an££iv (231, 304). In all instances where -££iv is followed by a vowel, the infinitive ending stands before the main caesura of the third foot. The Scutum thus faithfully replicates the Homeric state of affairs, which is not surprising given that the central part of this poem is modeled on the Homeric description of the shield of Achilles (E 478-608) and the Scutum is in fact, in G. Most's words, "a striking example of interaction between the Hesiodic and Homeric poetic traditions" (Most 2006: lv). In the pseudo-Hesiodic fragments (which for the most part cannot be dated with any certainty) we find three examples of thematic aorist infinitives in -££iv, all of which are found in verses that are either direct borrowings from the Homeric text or have been constructed from Homeric material.25
24 The date of the composition of the Scutum is traditionally placed somewhere between 590 and 570 BCE. The terminus ante quem is based on the name of the centaur Melanchaetes on the Francois vase, likely due to a misunderstanding of Sc. 186, as well as on the much-discussed indication found at the end of Hypothesis A to Scutum that Stesichorus (who probably floruit ca. 570-540) knew the poem. The terminus post quem can be set at about 590: in the lines 390401 the particular season in which Heracles faced off Cycnus is given as mid-July and the only reason for providing this information could have been the existence of a specific festival at Apollo's temple at Pagasae, where the action is said to take place. Such a festival could only have been instituted after the Sacred War when the Anthelan Amphictyony demolished the city of Crisa (591 BCE): the fall of Crisa was equated with Cycnus' demise (see Solmsen 1965: 3, Fn. 4; Janko 19864).
25 npa0ee, iv |aaAa nep fie^awxeg (fr. 35.3 Merkelbach-West) can be viewed as an imitation of I 532 5ianpa0eeiv fie^awxeg Ap^i (compare Sc. 240 npa0eeiv fie^awxeg). The second example is found
The Hesiodic situation in regard to thematic aorist active infinitives is thus different from what we find in Homer and post-Homeric hexameter poetry: only 18% of these forms in Homer end in -£iv, while 100% do in Hesiod.26 This fact has not been duly exploited in the studies of Hesiod's language and poetic technique. In view of the paucity of active thematic aorist infinitives attested in Hesiod it might not seem to be totally unreasonable to question the validity of statistics in this case, but in fact the form has just the same frequency as it has in Homer: 12 aorist infinitives in -£iv for 1813 lines27 gives us an average of one form in every 151 lines, while in Homer (27812 lines) there are 192 forms of active thematic aorist infinitive, viz. one form for every 145 lines.
Unfortunately, even though generally Hesiod is just as formulaic as Homer, in Theogony and Works and Days there is not a single formula, with or without Homeric parallels, that contains an infinitive of the type we are interested in (or any infinitive, in fact).28 Whether or not this gap is significant, is probably impossible to say,29 but as a result we cannot see in the same way as we can in the case of Homeric epics how exactly Hesiod may have handled some of the cases that must have come down to him from the stock of traditional material. Nevertheless, the conclusion that emerges is that even though the dialect of Hesiod's poems is «essentially the same as that
in fr. 58.12 L5e[eiv A]a^n.[pov ^aog qeAioio which can be compared to Homeric verse-end formulae C^ei Kai op^ ^aog qeAioio and o^ea0ai Aa^npov ^aog qeAioio?). Lastly, eK^uYeeiv Kai aAu^ai (fr. 150.29) may have come directly from 0 243 =^ 216.
26 In Homer there are 35 active infinitives of the thematic aorist ending in -elv, 55 forms in -e^ev and 102 forms in -eeiv.
27 The "Days" section of Erga omitted; for argumentation see e.g. Solmsen (1963), who bracketed the entire section in his OCT text.
28 As can easily be seen by consulting the collections by Krafft (1963) and Pavese (1972: 123-77).
29 Note in any event that a spondee in the fourth or the fifth foot is avoided in Hesiod even more assiduously than in Homer: see the figures in Paulson 1887: 4.
of Homer»30, the poet did not have access to the poetic tradition where thematic aorist infinitives in -££iv were freely available as substitutes for *-ehen in traditional formulae.31 This paper is not the place to put forth far-reaching theories concerning the formation and constituent parts of Hesiod's poetic language: my purpose here is merely to signal an important isogloss separating the Homeric and Hesiodic dialects. The problem merits further study.
Let's take stock. In this paper I have argued that under the traditional theory of an «Aeolic phase» in the development of Homeric epic diction it is hard (if not impossible) to adequately explain the distribution of active thematic aorist infinitives in Homer where we find Aeolic -£^£v if the preceding syllable is heavy (-,4/5w w), but Ionic (likely kunstsprachliches) -££iv before consonant or caesura if the preceding syllable is light (w w -,3/4/5): if -£^£v was the sole thematic active infinitive ending in Aeolic epic poetry, it is unclear why Ionic bards would replace a traditional form by an artificial one without any metrical reason to do so
44
(*paA£^£, v &£ > paA££, iv &£ ). But in case there never was a full-scale «Aeolic phase», things appear in an entirely different perspective: after the infinitive suffix *-ehen contracted to -en, the Ionic singers were nevertheless able to continue employing traditional formulae by replacing the aorist infinitive in former *-ehen (w w - before a consonant) by a likewise Ionic form in -££iv. In cases where *-ehen used to fill the biceps and a spondaic foot would be undesirable the bards had no choice but to borrow (or create) Aeolic forms in -£^£v. The absence of aorist infinitives in -££iv from Hesiod is likely to be significant:
30 West 1966: 79.
31 Despite the fact that there are uncontracted present infinitives of -ee/o- verbs in Hesiod, e.g. npoxeeiv (E. 596), viz. the very forms that served as the trigger for the creation of Homeric aor. inf. in -eeiv.
these forms seem to point to an Ionic poetic tradition alien to Hesiod.
The distribution of the forms of active thematic aorist infinitive thus supports the diffusionists' view of the Aeolisms in the Homeric dialect. It may not be the most important argument in favor of this view, but it is one.
References
Зайцев 1999 — Зайцев А. И. «Греческая метрика» Б. Снелля // Б. Снелль. Греческая метрика. М.: Греко-латинский кабинет, 1999. С. 5-6.
Казанский 2004 — Казанский Н. Н. «Микенская словесность» и проблема диалектной стратификации древнегреческой эпической традиции // Индоевропейское языкознание и классическая филология — VIII. Материалы чтений, посвященных памяти профессора И. М. Тронского. / Отв. ред. Н. Н. Казанский. - СПб.: Наука, 2004. С. 88-96.
Bechtel 1921 — Bechtel Fr. Die griechischen Dialekte. Bd. 1: Der lesbische, thessalische, bootische, arkadische und kyprische Dialekt. Berlin: Weidmannsche Buchhandlung, 1921.
Bekker 1863 — Bekker I. Homerische Blatter. Bonn: A. Marcus, 1863. Berg-Haug 2000 — Berg N., Haug D. Innovation vs. tradition in Homer — an overlooked piece of evidence // Symbolae Osloenses. 2000. Vol. 75. P. 5-23.
Blumel 1982 — Blumel W. Die aiolischen Dialekte: Phonologie und Morphologie der inschriftlichen Texte aus generativer Sicht. Gottingen: Vandenhoeck und Ruprecht, 1982.
Chantraine 1973 [1958] — Chantraine P. Grammaire homerique. 3eme tir. T. 1. Paris: Klincksieck, 1958.
Drerup 1903 — Drerup E. Die Anfange der hellenischen Kultur.
Homer. Munchen, Kirchheim, 1903.
Forssman 1991 — Forssman B. Schichten in der homerischen Sprache // Zweihundert Jahre Homer-Forschung / Hrsgg. von J. Latacz. Stuttgart; Leipzig: B. G. Teubner. S. 259-288.
Garcia-Ramon 1973 — Garcia-Ramon J. L. El llamado sustrato eolico: revision critica // Quadernos de filologia clasica. 1973. T. 5. P. 235-277.
Garcia-Ramon 1975 — Garcia-Ramon J. L.____________Les origins
postmyceniennes du groupe dialectal eolien: etude linguistique. Salamanca: Universidad de Salamanca, 1975.
Garcia-Ramon 1977 — Garcia-Ramon J. L. Le pretendu infinitif «occidental» du type exev vis-a-vis du mycenien e-ke-e // Minos. 1977. 16. P. 179-206.
Hackstein 2002 — Hackstein O. Die Sprachform der homerischen Epen: Faktoren morphologischer Variabilitat in literarischen Fruhformen: Tradition, Sprachwandel, Sprachliche Anachronismen. Wiesbaden: L. Reichert, 2002.
Hackstein 2010 — Hackstein O. The Greek of epic // A companion to the Ancient Greek language / Ed. by E. J. Bakker. Malden, MA: Wiley-Blackwell. P. 401-424.
Hajnal 2003 — Hajnal I. Troia aus sprachwissenschaftlicher Sicht: die Struktur einer Argumentation. Innsbruck: Institut fur
Sprachwissenschaft der Universitat Innsbruck, 2003.
Haug 2002 — Haug D. Les phases de l'evolution de la langue epique. Trois etudes de linguistique homerique. Gottingen: Vandenhoeck und Ruprecht, 2002.
Hermann 1914 — Hermann E. Die epische Zerdehnung // Zeitschrift fur vergleichende Sprachforschung ( = Kuhns Zeitschrift). 1914. Bd. 46. S. 241-265.
Hooker 1977 — Hooker J. T. The language and text of the Lesbian poets. Innsbruck: Institut fur Sprachwissenschaft der Universitat Innsbruck, 1977.
Horrocks 1987 — Horrocks G. C. The Ionian epic tradition: was there an Aeolic phase in its development? // Minos. 1987. T. 22-24 ( = Studies in Mycenaean and Classical Greek presented to John Chadwick / Ed. by J. T. Killen, J. L. Melena and J.- P. Olivier). P. 269-94.
Horrocks 1997 — Horrocks G. C. Homer's dialect // A new companion to Homer / Ed. by I. Morris and B. Powell. Leiden; New York: Brill. P. 193-217.
Janko 1982 — Janko R. Homer, Hesiod, and the Hymns: diachronic development in epic diction. Cambridge [UK.]; New York: Cambridge University Press, 1982.
Janko 1986 — Janko R. The shield of Herakles and the legend of Cycnus // Classical Quarterly. 1986. 36/1. P. 38-59.
Krafft 1963 — Krafft F. Vergleichende Untersuchungen zu Homer und Hesiod. Gottingen: Vandenhoeck and Ruprecht, 1963.
Mangold 1873 — Mangold B. De diectasi Homerica imprimis verborum in -aw // Studien zur griechischen und lateinischen Grammatik. 1873. Bd. 6. S. 139-213.
Meister 1921 — Meister K. Die homerische Kunstsprache. Leipzig: B. G. Teubner, 1921.
Meyer 1856 — Meyer L. Der infinitive der homerischen Sprache. Gottingen, 1856.
Miller 1982 — Miller D. G. Homer and the Ionian epic tradition. Innsbruck: Institut fur Sprachwissenschaft der Universitat Innsbruck, 1982.
Nagy 1974 — Nagy G. Comparative studies in Greek and Indic meter. Cambridge, MA, 1974.
Parry 1932 — Parry M. Studies in the epic technique of oral verse-making. II. The Homeric language as the language of an oral poetry // Harvard Studies in Classical Philology. 1932. Vol. 43. P. 1-50.
Paulson 1887 — Paulson J. Studia Hesiodea I: de re metrica. Lund: Malstrom.
Pavese 1972 — Pavese C. O. Tradizioni e generi poetici della grecia arcaica. Rome: Edizioni del'Ateneo, 1972.
Peters 1986 — Peters M. Zur Frage einer 'achaischen' Phase des griechischen Epos // O-o-pe-ro-si. Festschrift fur E. Risch zum 75. Geburtstag / Hrsgg. von A. Etter. Berlin; New York: W. de Gruyter, 1986. S. 303-319.
Peters 1989 — Peters M. Sprachliche Studien zum Fruhgriechischen: Habilitationschrift. Universitat Wien, 1989.
Renner 1868 — Renner J. G. De dialecto antiquioris Graecorum poesis elegiacae et iambicae // Studien zur griechischen und lateinischen Grammatik. 1868. Bd. 1. P. 133-236.
Ritschl 1838 — Ritschl F. Die Alexandrinischen Bibliotheken unter den ersten Ptolemaern und die Sammlung der Homerischen Gedichte durch Pisistratus nach Anleitung eines Plautinischen Scholions. Nebst litterarhistorischen Zugaben uber die Chronologie der alexandrinischen Bibliothekare, die Stichometrie der Alten, und die Grammatiker Heliodoru. Breslau: G. P. Aderholz, 1838.
Ruijgh 1995 — Ruijgh C. J. D'Homere aux origines proto-
myceniennes de la tradition epique. Analyse dialectologique du langage homerique, avec un excursus sur la creation de l'alphabet grec // Homeric Questions. Essays in Philology, Ancient History and Archaeology / Ed. J. P. Crielaard. Amsterdam, 1995. P. 1-96.
Rzach 1875-76 — Rzach A. Der Dialekt des Hesiodos // Jahrbucher fur classische Philologie. 1875-76. Suppl. Bd. 8. S. 355-466.
Schmidt 1968 — Schmidt V. Sprachliche Untersuchungen
zu Herondas. Berlin: De Gruyter, 1968.
Schwyzer 1939 — Schwyzer E. Griechische Grammatik auf der Grundlage von Karl Brugmanns Griechischer Grammatik. 1. Bd.: Lautlehre, Wortbildung, Flexion. Munchen: C. H. Beck, 1939.
Simmerle 1874 — Simmerle P. M. Zur Bildung des homerischen Infinitivformen. Innsbruck: Wagner'schen Buchdruckerei, 1874.
Smyth 1894 — Smyth H.W. The sounds and inflections of the Greek dialects. Ionic. Oxford: Clarendon Press, 1894.
Solmsen 1963 — Solmsen F. The "Days" of the Works and Days // Transactions and Proceedings of the American Philological Association. 1963. Vol. 94. P. 293-320
Solmsen 1965 — Solmsen F. Ilias XVIII, 535-540 // Hermes. 1965. 93. P. 1-6.
Thumb-Scherer 1959 — Thumb A. Handbuch der griechischen Dialekte. Bd. 2 / 2. erw. Auflage von A. Scherer. Heidelberg: C. Winter, 1959.
Wathelet 1970 — Wathelet P. Les traits eoliens dans la langue de l'epopee grecque. Roma, Edizioni dell'Ateneo: 1970.
West 1966 — Hesiod. Theogony / Ed. with prolegomena and commentary by M. L. West. Oxford: Clarendon Press, 1966.
West 1982 — West M. L. Greek metre. Oxford: Clarendon Press, 1982.
West 1988 — West M. L. The rise of the Greek epic // Journal of Hellenic Studies. 1988. Vol. 108. 151-172.
West 1998 — West M. L. Grated cheese fit for heroes // Journal of Hellenic Studies. 1998. Vol. 118. P. 190-191.
Witte 1915 — Witte K. Wortrhythmus bei Homer // Rheinisches Museum fur Philologie. 1915. Bd. 70. S. 481-523.
Witte 1972 — Witte K. Zur homerischen Sprache. Darmstadt: Wissenschaftliche Buchgesellschaft, 1972.
Wuest 1943 — Wuest H. Morphologie des homerischen Infinitivs auf Grundlage der Ablautserscheinungen. Dissertation Freiburg i.d. Schweiz. Freiburg in der Schweiz: Paulusdruckerei, 1943.
Wyatt 1992 — Wyatt Jr. W. F. Homer's linguistic forebears //
Journal of Hellenic Studies. 1992. Vol. 112. P. 167-173.
