CC BY
33
Резюме. В статье разбирается употребление др.-гр. duauooc и (а)даирою, для которых вслед за П. Маасом (и вопреки LSJ) предлагаются значения ‘слабый, ничтожный, бессильный’ и ‘ослаблять, уничтожать’. Основу ацаиро- предлагается возводить к композиту *n-meh2ur-o- ‘не имеющий роста / силы’, вторая часть которого является отглагольным именем гетероклитического склонения, образованным от глагольной основы *meh2(i)- ‘становиться большим; быть большим’: хетт. mai- ‘расти; процветать’, слав. *matoru-, *materu-, лат. maturus, а также, согласно остроумной гипотезе И. С. Якубовича, праиран. *maH- ‘быть, становиться’. В конечном итоге все эти формы восходят к корню *meh2- ‘большой’ (др.-ирл. mar, гот. mais).
Ключевые слова: этимология, лексикология, древнегреческий язык, хеттский язык, иранские языки, система Каланда.
1. According to the LSJ9, the meaning of a^aupo^ is ‘hardly seen, dim, faint’; this lemma has not been substantially revised in the 1996 Supplement and the same meaning appears in a number of other dictionaries1. There is little doubt that the word came to mean something like ‘dark’ already in the Byzantine period; the question is what it meant in antiquity.
One major work of reference where a different opinion is stated is W. Cronert’s ambitious but unfinished revision of F. Passow’s Handworterbuch der griechischen Sprache (itself the basis for the original LSJ): there in the entry penned by P. Maas a^aupo^ is glossed as ‘schwach, matt, trube’2. In what follows I will attempt to
* It is a pleasure to thank Jay Jasanoff, Martin Peters, and Jeremy Rau for their comments on earlier versions of this paper. Usual disclaimer applies. I would also like to gratefully acknowledge the financial support from the Research Council of the President of the Russian Federation (grant nr. MK-389.2011.6).
1 The DGE, for instance, gives ‘dim’ as the basic meaning (“tenue, borroso”), while offering ‘weak’ (“flojo, debil”) as secondary.
2 Cronert 1912: 342. Maas probably got this idea from his teacher Wilamowitz, according to whom ‘schwach’ was the “right” meaning of the word (Wilamowitz-Moellendorff 1895: 35).
demonstrate that Maas was right and the sense ‘dark, dim’ is a secondary development from the original meaning ‘weak, feeble’; once the correct meaning is established, a novel etymology for the word will be proposed3.
1. 1. The two earliest attestations of this epithet are in the Odyssey and in Sappho, in contexts that are sufficiently close to each other to be examined jointly. In both passages a^aupoq refers to phantoms: in the Odyssey it is the ciSwAov, a vision of Penelope’s sister Iphthime, while in Sappho the epithet is applied to vskus^, ghosts in Hades:
Od. 4.824 = 835
x^v S’ аладегводеуоу лроаефл eiSroAov dp,aupov The dp,aupo<; phantom said to her in reply
Sappho 55 Voigt
KaxOavoiaa Ss Ksiani otiSs пота ^vap-oawa asOsv saasx’ otiSs |пок’| uaTspov oti yap nsSsxni? PpoSrov T®v sk nispia^ aAX’ афауп<; Kav AiSa Soproi фOlтdanl? nsS’ dpauprov vsKurov SKnsnoTapsva.
But when you die you will lie there, and afterwards there will never be any recollection of you4 since you have no share in the roses of Pieria; unseen' in the house of Hades, flown from our midst, you will go to and fro among the a^aupo- dead.
Despite the difference between the two passages — a description of a dream in Homer, a threat of withholding immortality granted through poetry in Sappho5 — it comes as no surprise that that the
3
I am not aware of any lexicographic studies of аряирос; besides McKinlay 1957 which is a somewhat idiosyncratic and disorganized collection of all passages where аряирОс; is used listed together with translations proposed for them to date.
4 The text in v. 2 is uncertain: Bucherer (1904: 106) printed ou§£ пОва slc; UGispov “or any longing” (adopted by Wilamowitz 1913: 88 n. 2 and Campbell 1982; for a new recent argument in favor of this emendation see Clay 1993 who noted the alliterative sound play: пота - пОва -EKnsrcoTapiva); Page 1955: 137 suggested оиб’ La tolc; uaTspov.
5 The interpretation of the poem, especially the last two and a half lines, is not universally agreed upon, see Page 1955: 137. The problem is caused not so much by the text itself, as by Plutarch’s testimony: he cites the first two lines twice, saying in one place that the poem was addressed to an uneducated woman and in another place that it was addressed to a rich woman. It is therefore uncertain whether the poem is personal at all.
same epithet is used to characterize the apparition of a living person (Iphthime) and the ghosts of the dead: the Greeks called both siSro^a. We need only compare Sappho’s image of the aimlessly wandering dead with the words of Patroclus’ spirit visiting Achilles:
II. 23.72-74
xqke pe e’Lpyouoi yuxa'i £i5wXa Ka^ovxwv, ou§£ pe n© piayeaOai unep noTapolo ewaiv, akk’ arn©c; akaknpai av’ eupunukec; AiSoc; 5w.
The souls, the images of dead men, hold me at a distance, and will not let me cross the river and mingle among them, but I wander as I am by Hades’ house of the wide gates6.
How were these siSro^a imagined? Shadow, of course, is one major recurrent motif, found, for instance, in the scene from the Nekuia where Odysseus tries in vain to embrace the dead ghost of his mother which flows from his hands like a shadow, oki^ slks^ov (Od.
11.207), or Pindar’s famous oKid<; ovap ‘dream of shade’ (P. 8.95) referring to the hero Amphiaraus visualizing his son Alcmaeon (see Nagy 2000: 110-13). Based on the idea of the shade, a^aupov siSro^ov in the Odyssey has been taken to mean a ‘dark, shadowy spectre’ and a^auprov vsKurov in Sappho has been understood as ‘shadowy dead’. Further variations on the same notion have included ‘dim’, ‘dark’, ‘faint’ and ‘obscure’7.
However, connotations of shadow include not only visual effects, but also weakness and insignificance. For instance, the siSro^ov of Patroclus escapes from Achilles like vapor, ^uxs Kanvo^ (Il. 23.100): this comparison emphasizes lack of physical substance more than anything else. It behooves us therefore to consider an alternative interpretation of a^aupov siSro^ov as ‘frail phantom’.
Among the epithets of phantoms and ghosts in early Greek poetry a particularly frequent one is a^svnvo^, usually translated as ‘helpless’, e.g. rcu^ai a^svnvwv ovsLprov ‘gates of helpless dreams’
6 The expression ei5©ka KapovT©v is also used at Od. 11.476; 24.14.
7 For instance, A. Luppino (1967: 290) has suggested ‘oscuri morti’, emphasizing the idea of “non-visibility” expressed earlier in the same sentence by a^avn^ as well as by A[§n^ (*n-uid-), but see the cogent objections by E. Tzamali (1996: 284-85); in particular, Tzamali has presented an interesting argument in favor of understanding a^avn^ here as ‘fame-less’ (cf. ou§£ pvapoouva). However, Tzamali’s own interpretation of apaup©v veKU©v as ‘unbedeutenden Toten’, inspired by Hesiod’s use of the word in the sense ‘demoted’, in my opinion lacks conviction.
(Od. 19.562) or vsKurov d^svnva Kapnva ‘helpless heads of the dead’ (4x Od.)8. That the word indeed means ‘without power’ is clear from its other uses: in the Iliad it refers to hypothetical consequences of a wound received by Ares and in Sophocles’ Ajax it refers to the mental impairment of the hero:
II. 5.885-87
q x£ ks §npov
5 ~ / % V Ч ) /ГЧ
auxou rcqpar enaaxov ev aLvqaiv vsraosaaiv, ц ksv a^sv^vOq ea xaAxoio xun^ai.
Otherwise I should long be lying there in pain among the stark dead men, or go living without strength because of the strokes of the bronze spear
Soph. Aj. 887-90
oxsxXia yap eps ys xov paKpwv dXarnv novrov оирш p^ nsXaaai 5popw,
\ V о» \AA <г/
dkk a^sv^vov avopa p^ Asuaasiv onou.
It is cruel
that I, who have roamed with such great toil, cannot come near him with a fair course, but fail to see where the enfeebled man is.
To sum up the argument thus far, the adjective d^aupoc in its two earliest attestations is not any likelier to mean ‘dark’ than ‘weak, feeble’. The similarity in use between d^aupoc and d^svnvoc is potentially instructive, since the latter is universally agreed to have had the original meaning ‘weak’9.
Incidentally, in Aristophanes d^svnvoc stands back to back with the compound d^aupoPioc ‘having d^aupo- life’ in the introduction to the bird cosmogony. This rich passage turns out to be quite informative for our purposes:
Ar. Av. 685-88
V C* \ /■ v C* "> r П ' Л Л '
ays 9^aiv avopsc a^aupopioi, 9uAA®v ysvsa npoaopoioi,
oXiyoSpavssc;, nMaparn nn^ou, aKiosiSsa фОХ’ dpev^va, drcrqvsc; ^npepioi xaXaoi Ppoxoi dvspec; sLKsXovsipoi, npoasxsxs xov vouv tolc dOavaxoi^ ^piv...
8
According to Jebb (1907), the notion here is ‘unsubstantial’.
9 apevnvoc; is a privative compound with pevoc; in its archaic meaning ‘strength’ used as the second member; the word is either extended with a further suffix *-e/ano- (see Schwyzer 1939: 490) or, more likely, modeled after aKpnvoc (Frisk 1960-72: 91; Beekes 2010: 86).
Come now, you men out there, who live such dpaupo- lives— you’re frail, just like a race of leaves—you’re shaped from clay, you tribes of insubstantial shadows without wings, you creatures of a day, unhappy mortal men, you figures from a dream, now turn your minds to us
The chorus of birds clearly emphasizes the fragile, transcendent nature of the human race; the same idea of temporality is evoked by the comparison of men to leaves, echoing Homer and Mimnermus (Il. 6.146; Mimn. fr. 2.1-8 W.2). The adjective OKiosiSsa may have prompted the translation ‘living in darkness’ adopted by the LSJ, but the general idea of frailty which this passage focalizes is much better captured by N. Dunbar’s ‘living feeble lives’ (Dunbar 1998: 391).
Let us now briefly comment on the use of a^avpoq in tragedy and propose a scenario for the subsequent development of the semantics of the word. Thanks to its associations with the spirits of the dead, the word has evolved into a generalized sinister epithet:
Aesch. Choe. 157
kMs Se poi ae^a^, kXu’, © Seanox’, e£, / apaupac ^pevoc hear me, oh hear me, my honored lord, out of your apaupo- mind
A. Garvie has correctly observed that the ^p^v is that of deceased Agamemnon (Garvie 1986: 84). Garvie’s comparison with %0ov(a 9pcv( (Pi. P. 5.101) is likewise very compelling: just as in Pindar the sacred kings of older times hear with their minds beneath the earth of the great achievements, in Aeschylus a^aupoq simply signifies the fact that Agamemnon is now a ghost in Hades (a^aup^ meaning ‘of a^a'upwv vckuwv’)10. Such bold usages of a^aupoq must have given rise to reinterpretations of the obscure word: apparently it is in fifth century tragic poetry that the meaning ‘dim’, ‘shadowy’, ‘dark’ begins to emerge; in certain passages it has to be recognized as predominant (see e.g. Jebb 1907: 40 (ad Aj. 182) or Fraenkel 1950: 236 (ad Ag. 466)).
The use of a^aupoq by Sophocles, especially in Oedipus at
10 H. Friis Johansen and E. Whittle have included this passage among the evidence amassed by them for the heart or 9pev£<; or anlay^va to turn black from grief, fear or anger (1980, 3: 134-35); this interpretation, however, is unwarranted, in my opinion, since even the proponents of such interpretations of dpaupo^ as ‘dim’ or ‘dark’ have never regarded the word as a color adjective.
Colonus must have engendered particular confusion: many commentators have attempted to render a^aupw kw^w (OC 182) or a^aupai^ xsPc™ (OC 1639) as ‘blind steps’ or ‘blind hands’, assuming a metonymical transfer of Oedipus’ blindness to his limbs. But this does not seem necessary, and the meaning ‘weak, feeble’ would suit all passages just as well11.
The adjective a^aupo^ is thus closely associated with the dead, the otherworldly and the impaired.
1. 2. Next we should turn to the factitive verb a^auporo: does it mean ‘make dim’ or ‘make feeble’? The immediate answer is neither: the actual meaning of a^auporo in early Greek poetry is ‘to destroy’, ‘to make naught’. For instance, Pindar describes Perseus’ killing Medousa and Graiai in the following way:
Pi. P. 12.13
^xoi to xe Oeaneaiov QopKoi’ d^aupwosv yevoc; he destroyed the awesome race of Phorkos
The tragedians are fond of this verb which they likewise use in the meaning ‘to destroy’, although not quite so literally as Pindar:
Aesch. Ag. 464-66
\ V 1 V C* /
xuxnpov ovx’ aveu oiKac naXivxuxeL xpipdi piou xiOeia’ d^aupov
when a man is prosperous without justice, by wearing away his life in a reverse of fortune,
(the Erinyes) bring him to ruin
Eur. Hipp. 816
xlc dpa adv, xdXaiv’, d^aupoi Zoav;
Who brings your life to ruin, poor woman?
While the verb a^auporo applied to persons may be interpreted as ‘to make into a faint shade’ (viz. a ghost, a shadow roaming in Hades), this interpretation is quite unlikely in the passage from Euripides above where the verb is construed with Zoav (translated by W. S. Barrett as ‘brings to naught’) (Barrett 1964: 319).
The causal relationship between a^auporo and Slk^ (or, rather, lack thereof) found in the passage from Aeschylus above may hark
11 With dpaupw k©Xw Wilamowitz compared no5oc dpaupov ixvoq “feeble steps” (Eur. HF 123), Wilamowitz-Moellendorff 1895: 35.
back to Hesiod where the verb is applied to the consequences of unjust behavior:
Hes. Op. 323-25 (about a thief):
T S O ' / <"•» / 1 y /
eux’ av K£poo^ voov eqanax^a^
avOpwnrov, ai5w 5e x’ avaiSeLn KaxonaZ'fl,
Q8ia §£ piv ^aupouai 080L, pivuGouai §£ olkov When lust for profit deceives the intelligence given to mankind and when immodesty shamelessly casts out modesty wholly, readily heaven paupO-es that man and diminishes his house
. \ry
In this passage ^auporo is paired with ^ivu0ro, which may suggest that the former verb here has a meaning like “belittle”. Such an assumption can be countered by a comparison to Solon’s elegy:
Solon 4.34-37 W.2:
xpaxea ^8iatv8i, nau8i KOpov, uppiv a^aupoi, auaLv8i 5’ ax^c; av08a ^uOp8va 8u0uv8i 5e OLrac; aKoXiac;, un^p^ava x’ epya npauv8r nau8i 5’ epya OixoaxaaLn^
she ( = Euvopvn) smooths the rough, stops insolence, impairs violence,
she withers the budding flowers of ruin,
she straightens crooked judgments, haughty deeds
she assuages, and she stops the deeds of sedition.
V. 36 8u0uv8i 5e §LKa^ CKo^id^ with its distinctive Hesiodic feel makes it particularly plausible that Solon was relying on a Hesiodic didactic tradition here. But it is also clear that uPpi<; in v. 34 is not “diminished”: a stronger verbal action is called for. If — but it is only if — Solon was composing his elegy with Hesiod’s verses about divine justice in mind, it may be possible to link his use of a^aupOro with the Hesiodic and translate the verb in both poems by ‘destroy, bring to naught’.
1. 3. To sum up the results of our inquiry into the semantics of a^aupO^: it appears quite plausible that the adjective originally referred to physical weakness, death and lack of substance, and that the factitive verb a^aupOro meant ‘bring to naught, destroy’.
2. Having established the original meaning of a^aupO^, we can now revisit the etymology of the word. The handbooks offer virtually nothing beyond the customary “unsicher” and “inconnue”; still, there
12
The unusual form of the root will be discussed presently.
are a few suggestions on the record that should be discussed before a novel idea can be proposed.
2. 1. R. S. P. Beekes, well-known for his propensity for substrate etymologies, labeled d^aupo<; as Pre-Greek in his recent etymological dictionary (see Beekes 2010: 84). His main reason for choosing this approach in this particular case was the alternation between d^aupo- and ^aupo-, ^aupos/o-, such vacillation of the prothetic vowel allegedly being a hallmark of borrowings from the substrate language of the Balkans. However, there is another explanation for this alternation which appears more plausible (see Troxler 1964: 124): ^aupos/o-, limited to poetic language13, could be a product of a Leumann-style false resegmentation, whereby the initial vowel was reanalyzed as the final vowel of the preceding word14. A possible starting point could be provided by a verse like Hes. Op. 693 a^ova Kaua^ai^ Kai Qopxia uaupfflQsm (this word division is guaranteed metrically: ^opxt’ d^aupro08Ln would create a caesura after fourth trochee)1 . Spreading from such verses the consonant-initial form ^aupos/o-, ^aupo- gained terrain in the poetic language.
2. 2. A. Fick compared a^aupo^ to Germanic *maraa- ‘worn out’ reflected in Old English mearu, Old High German marawi and muruwi, Modern German murbe, and eventually going back to PIE root *merh2- ‘crush, wear out’ (Old Norse merja, Vedic mrnati, Greek ^apaLvro, Hittite marra- / marri-)16,17. This connection, based on a non-existing rule of PIE “u-epenthesis” (*-ra- > *-ura-)18, is of
13
Boisacq’s “Arcadien paupoc” is apparently a mistake: I have not been able to locate the word in the corpus of Arcadian inscriptions and glosses.
14 But not through a phonological loss of the initial vowel, despite Stromberg 1944: 44, cited with approval in the DELG and the LfgrE.
15 See Nauck 1872: 223 (= 1874: 269).
16 See Fick 1874: I, 718. The lack of initial vowel in Greek papaLv© is one of several things that make this connection very doubtful.
17 Despite Bjorvand-Lindeman 2007: 769 and Matasovic 2009: 268, Old Irish meirb and Welsh merw belong not to *merh2- ‘crush’, but to *mer-‘die’ (whatever ultimate relationship there may be between these roots, if any): the oldest meaning of meirb is ‘corpse’ (e.g. Ml. 113 b 8 inna merbi .i. mortua corpora), while the meaning ‘weak, feeble’ appears later.
18 An ardent proponent of this sound law was P. Kretschmer who in his long study of the “u-epenthesis” enthusiastically accepted Fick’s etymology (1892: 448).
course impossible phonologically: a protoform *amaruo- (however one may wish to derive it from *merh2-) would have given Ionic and Doric *a^apo^, Attic and Aeolic *а^аро<;19.
2. 3. Somewhat similar to Fick’s (and likewise erroneous) is U. von Wilamowitz’s inner-Greek explanation of a^aupo^ as an “aeolische nebenform zu amalos”, viz. as a product of an Aeolic development of *amaruo-20. This theory is easily countered by such examples as *arua ‘curse’ > Lesbian apa ^ - (Sapph. fr. 86.5) or *koraa ‘maiden’ > Lesbian кора ^ - (Sapph. fr. 53), Boeotian кора (IG 7.587, 6th cent.) which prove beyond any doubt that post-rhotic *u was eliminated in Aeolic without compensatory effects of any kind.
19 North-Germanic has a slightly different adjective with nearly the same meaning: Old Norse meyrr, Old Danish m0r, Old Swedish m0r, Faroese moyrur 1. ‘tender, easy to chew’ (of meat), 2. ‘tired, exhausted’. These forms go back to *maurija-. P. Persson was the first to suggest that this stem is formally comparable to Greek dpaupoc; (Persson 1892; adopted by Noreen 1894: 217, Zupitza 1899: 101; Cop 1956: 230). Persson went on to connect Germanic *maurija- (and, by extension, dpaupoc;) to *maraa-‘worn out’ mentioned above, using the same “u-epenthesis” as Fick before him. Let us try to clear this up:
Since a) the meanings of *mauriia- and *marua- are identical and b) these two stems are in complementary distribution in Germanic, the idea that they go back to the same prototype is in fact quite compelling. Even though it is not easy to cite examples of a regular Germanic metathesis that would convert *maraija- to *maurija-, the latter is customarily identified with *maraa- and PIE *merh2- ‘crush’ (Old Norse merja, etc.) in modern reference works (e.g. Heidermanns 1993: 404; Bjorvand-Lindeman 2007: 769 prefer to connect *maruiia- > *mauriia- specifically to Hittite marriye/a-tta, marra-tta ‘cook until tender’, emphasizing the semantic match). If this approach is correct, there would be no relationship with dpaupo^ (note the absence of initial h in Hittite marriye/a-).
If, however, Germanic *maurija- is not related to *maraa-, the comparison with dpaupo^ becomes a possibility to be considered seriously. But the semantic match is not ideal: the Scandinavian descendants of *mauri|a- all mean ‘tender’ or ‘tired’, while the meaning ‘decrepit’, cited by Persson, is attested in Old Norse only for the dvandva compound lasmeyrr, whose first part is a short form of adj. lasinn ‘weak, ill’ (all remaining attestations of meyrr refer to softness of some kind). Note that there is an alternative explanation of *maurna-: I. Lundahl (1930: 27) suggested that this word is an inner-Germanic derivative from a verb *mauen (Old Norse ma, pres. mai, pret. madi ‘wear out’, sagmugg ‘sawdust’); the root *meu- is taken to be a variant of protean *smeu- / *smei- by Falk-Torp 1960: 700-1 and de Vries 1962: 374.
20 Wilamowitz-Moellendorff 1895: 35.
2. 4. The last etymology to be reviewed was proposed by E. Zupitza (1899: 101) who in a brief note compared d^aupOt; to Russian (s)muryj ‘sullen, dark-grey’. This comparison found immediate acceptance in Prellwitz’s and Vasmer’s etymological dictionaries and continues to be cited in reference literature on Slavic, but not on Greek21. At first, it looks quite compelling: Russian (s)muryj, whose
-u- is a legitimate reflex of an earlier *-au-, can be reconciled with
22
Greek a^aupo^ under a reconstruction *(s)h2meh2uro- . It is true that the meaning ‘sullen, dark-grey’ of (s)muryj corresponds to what we have just decided was not the oldest meaning in Greek, but it would be methodologically unwarranted to discard the comparison on these grounds alone. However, Zupitza’s comparison is not free from other problems; in order to appreciate these, we need to examine the entire group of Slavic cognates.
Russian (s)muryj ‘sullen, dark-grey’ is related to Czech moura ‘grey or brown cow’, Slovene mur ‘black horse’ and Serbo-Croatian Mura attested as a name for dogs and goats23. It is generally agreed that this color adjective cannot be separated from the noun denoting black and dirty substances: Czech mour ‘soot’, Slovak mur ‘soot’, Serbo-Croatian mur ‘drift sand’, mura ‘mud, ooze; cement mortar; fertile soil’, Russian mur ‘mold’24. There are cognates in Baltic: Lithuanian mauras / mauras ‘mud, ooze, duckweed’ and Latvian maurs ‘lawn’.
The most plausible way of accommodating these data would be to reconstruct two stems: an adjective ‘dark, black’ and its abstract ‘dark, dirty stuff’ (whence ‘soot’, ‘mud’, etc.). Accent shift was one of the ways of making an adjective abstract in Proto-Indo-European and an accentual contrast between ‘dark’ and ‘dark stuff’ is therefore entirely expectable from the PIE point of view, cf. Vedic krsna-‘black’ : krsna- ‘antelope’ (*‘the black one’), Greek ^sukO<; ‘white’ : ^suko^ ‘white one’ (name of a fish). Such an accentual contrast would help to explain the puzzling intonational difference between the Slavic and the Baltic reflexes in our case: Latvian maurs points to acute immobile and while Lithuanian mauras / mauras is attested in all four paradigms, it is likely that the acute tone reflects the original
21 Prellwitz 1905: 31; Vasmer 1950-58, 3: 15.
22 On word-initial clusters of such structure see Nikolaev 2009.
23 Detailed documentation in Petleva 1973: 49 n. 58; ESSJa 20: 192.
24 For semantics compare Greek doi^ ‘mud, ooze’ vs. Hittite hanzana-‘black’ (< *h2(o)ms-no-), Sanskrit asita- ‘black’ (Nikolaev 2010: 239-42).
situation (see Derksen 1996: 86). Slavic *(s)murb, however, belonged to the mobile paradigm (a.p. c) and, accordingly, must have been oxytone in Balto-Slavic times . We thus have a pair *(s)m'auras and *(s)maur'as which can be traced back to an earlier pair of oxytone adjective *(s)h2meh2ur6- ‘dark’ and its barytone abstract *(s)h2meh2uro- ‘dark stuff’26.
*(s)h2meh2ur6- ‘dark’ would indeed have given both a^aupo^ and Russian (s)muryj, as well as the other Balto-Slavic forms mentioned above. But there are several problems with this reconstruction. First, adj. *(s)h2meh2ur6- with its full-grade in the root looks a little suspect and one would imagine that this full grade is a result of an analogical transfer from the substantive27. Since Greek -au- is best explained from *-ah2u- (< *-eh2u-), it seems reasonable to assume that the zero grade of *(s)h2meh2uro- would show a metathesized *-uh2- > *-u-. But Lithuanian muras ‘soft soil, mud’ with a short vowel in the root effectively excludes *(s)h2muh2ro- with a root-internal *-h2-.28 We are therefore left with proto-forms *(s)h2mauro- and *(s)h2muro- with an underlying *a: while not impossible, this reconstruction looks less compelling.
Secondly, Zupitza’s etymology of a^aupo^ is rendered problematic by the Armenian material, left unmentioned in the treatments of Balto-Slavic *mauras. H. Petersson compared Russian (s)muryj with Classical Armenian mor, -i ‘dirt, mud, mire, slime,
25
Cf. Proto-Slavic *mig'L (a.p. c) = Sanskrit megha- ‘cloud’, *svet& ‘light’ = Sanskrit sveta- ‘light’, *mext ‘bag; fur’ = Sanskrit mesa- ‘ram’, *xodt ‘move’ = Greek oboe, ‘way’, *golst ‘voice’ = Germanic *kalzan (Old Norse kal) ‘call’.
26 *-VHu- sequences being always realized as *-VHu- in Balto-Slavic, one might expect Hirt’s Law to produce *(s)maur'as > *(s)m'auras > *mur (cf. *dhuh2m6s ‘smoke’ > *dum'as > *d'umas > Slavic *dymt, Serbo-Croatian dim). This failure of Hirt’s Law to apply in an oxytone formation is exactly comparable to other high-profile exceptions to this law: Lithuanian gyvas, fem. gyva, Slavic *zTvl, Serbo-Croatian ziv ‘living’ (< *gwih3u6-) or Lithuanian sunus, acc. sg. sun^, Slavic *synt, Serbo-Croatian sin ‘son’ (< *suh2nu-). A new explanation of these troubling facts has been offered in Jasanoff 2008: 353.
27 On the problem of full-grade ro-adjectives see Vine 2002.
28 Old Irish mur ‘mire, shoal’ (LU 67 b 16; LL 58 a 9) can in theory go back to *(s)(hx)muhx-ro-, but Stokes (1901: 470) is probably right that this word is a Germanic borrowing, either from Old Norse myrr ‘slime’ or from Old English myre, English mire.
9Q
puddle, fen’ and o-stem mrur, gen. mrroy ‘sediment’ . In view of the near-perfect semantic match with the Balto-Slavic material, this comparison looks very plausible and Petersson’s reconstruction *mauri- is unassailable. But the lack of a prothetic vowel in mor and mrur is another nail in the coffin of Zupitza’s etymology, since in a reflex of PIE *h2mu- one would expect to find an initial a- in Armenian (cf. *h2ner > ayr ‘man’, *h2reu- > arew ‘sun’, hineh3mn > anun, etc.) 30.
We have to conclude that there is no connection between Greek a^aupo^ and the Baltic, Slavic and Armenian words just discussed31.
3. It appears that all previous attempts to solve the mystery of d^aupo- have not been successful. A new solution is therefore called for.
3. 1. Under the premise that d^aupo- is inherited, it can only continue one of the two following protoforms: *nmeh2uro- or *h2meh2uro-. The latter reconstruction does not seem to lead anywhere, but the former protoform, if segmented as *n-meh2ur-o-, is immediately analyzable as a possessive compound whose second member is derived from a heteroclitic stem *meh2ur / -n-, cf. *uodr /
" /-s o
-n- ^ *n-udr-o- ‘waterless’ (Gk. avu§po^, Skt. anudra-). This analysis corresponds nicely to what is otherwise known about the
29 Petersson 1916a: 42-43; 1916b: 280-81. For attestations see Acharyan 1971-1979, 3: 373-74. The word is not attested in Biblican Armenian (where mawr always stands for gen.sg. of mayr ‘mother’); in Modern Eastern Armenian the descendant of this word is mur ‘soot’. Curiously, the word is not included in the recent etymological dictionary by Martirosyan (2010).
0 Initial *h2- would have left no reflex in Armenian mOr thanks to Saussure-effect (if mor went back to *(s)h2moh2uro-), but it would have stayed in the precursor of mrur, since the latter cannot continue a protoform with *-au- ( < *-ou-) in the initial syllable.
31 It may be possible to set up an adjective *mur6- meaning ‘black’ (hence ‘dirty’) and its abstract *m6uro- ‘dirt, black stuff’ (whence adj. *mour6-through ablaut leveling). Further derivatives of *meu-, *mu- ‘black’ may include Russian mul ‘ooze’ (Vasmer 1950-58, 3: 10-11). Despite Lundahl 1931 and Pokorny 1951: 742, I doubt that these words have anything to do with the PIE root reflected in OCS myti some derivatives of which seem to have denoted liquid substances, in particular, filthy ones: Sanskrit mutra-‘urine’, Avestan muOra- ‘excrements’, Czech mydlo ‘soap’ < *muhxtlo-, Old Irish mun ‘urine’ < *muhxno-. These words clearly show that *meuhx-had a final laryngeal.
behavior of heteroclitic nouns in composition, namely the observation that their “externally” derived second compound members use the -r-stem (see Weiss 1994: 95). The presence of a distinct feminine form aiaaupa (Aesch. +) does not stand in the way of analyzing a^aupog as a compound: the adjective must have been sufficiently obscure in order to be reanalyzed as a simplex and thus capable of forming a feminine.
The next obvious step would be to identify *-meh2ur- as the strong stem of *meh2ur / -n-, a *-ur /-uen-stem made from a root *meh2-. It remains to find the root. None of the known roots of this shape provides a compelling semantic solution. Let us list the available options:
32
• *meh2- ‘to notify’ (Gk. (Dor.) pavu©, Lith. moju) ;
• *meh2- ‘to happen at the proper time’ (Hitt. mehur ‘time’, Lat. maturus)33;
• *meh2- ‘to want, to be eager’ (pa(i)opai, Elean opt. aor. med. paixo,
34
perf. pepaa, intens. paipa© ‘to quiver with eagerness’)
3.2. Barring some genuinely far-fetched semantic or formal scenarios, the only way out of the impasse seems to be lying outside the known roots. If the oldest recoverable meaning of a^aupo^ is ‘faint, weak, dead’, it is fair to assume that the underlying root *meh2- must have had a meaning ‘to be strong, to be alive’. Evidence for just such a root has recently become available.
Middle and Modern Iranian languages have preserved a group of cognates with the meaning ‘to become, to be’: Khotanese hama- ‘to become, to be’ (< *fra-ma-), Sogdian pret. (w)mt ‘to be’ (< *ui-ma-), Wakhi bimuy- ‘to be’ (< *a-ma-), etc. On the basis of these forms J. Cheung (2007: 257) has plausibly reconstructed a Proto-Iranian root *maH- ‘to become, to be’, to which I. Yakubovich (2010: 479) has added Sogd. "mt(')yc ‘true’ (< *a-ma-, cf. Ved. satyam, Engl. sooth for semantics).
While Cheung did not offer an etymological connection outside Iranian for the root that he posited, Yakubovich took a step further by comparing Iranian *maH- with the Hittite verb mai- / miya--1 ‘to grow, to thrive, to prosper’, mid. miyari ‘is born’, miyatar ‘growth’,
32 2
To the evidence for *meh2- ‘to give sign’ assembled in LIV 425 one might perhaps add Avestan a-ma(ii)- ‘show’, usually referred to *meh1-‘measure’.
33 Not in LIV. See Eichner 1973; Fortson 2007.
34 Not in LIV.
etc. The morphology of the verb is not in doubt: clearly we are dealing with i-present35, made from a laryngeal-final root36. Opinions, however, have been divided regarding the verb’s origin. Basically there have been two schools of thought: one, going back to Oettinger (1979: 471) and codified in LIV2 428, views the meaning ‘to be born’ as the original one and takes Hittite mai- from *meih1/3-‘to ripen’ with further connections to Hittite miu- ‘soft’, Latin mitis ‘mild’, Sanskrit mayas- ‘refreshment, enjoyment, pleasure, delight’ and Lith. mi'elas ‘pleasant’. The absence of other evidence for verbal *meih1/3-, the lack of any semantic connection between mai- and miu- in Hittite and the extreme semantic vagueness of other proposed connections all render this etymology not very likely. An alternative view was put forth by E. Forrer, who emphasized the idea of growth and connected Hittite mai- to Old Irish mar ‘big’, Gothic mais ‘more’37.
Yakubovich rejected both views and instead proposed a reconstruction of a new root *meh2- ‘to grow’, also reflected in Slavic *matoru- / *materu- ‘full grown’ and Latin maturus ‘id.’.38 Hittite, Slavic and Latin forms unequivocally point to ‘grow’, not ‘be’, as the basic meaning of *meh2-, while Iranian *maH- (exclusively combined with preverbs) probably started life with some more concrete meaning, like ‘grow up/out’.
3. 3. While the comparison between Iranian *maH- and Hittite mai- strikes me as plausible and very likely to be correct, there is one aspect of Yakubovich’ theory that I would like to revise. While the problems with the etymology of Hittite mai- adopted in the LIV (*meih1/3-) are obvious, the derivation from *meh2- ‘much’ chosen by Pokorny and Kloekhorst is in fact quite plausible: it seems to me that Yakubovich dealt with the latter proposal in a somewhat cavalier fashion. According to Yakubovich, — who operates with a nonstandard version of the laryngeal theory — the problem with deriving mai- from *meh2- is that Old Irish mar ‘great’ and Gothic mais ‘more’ cannot go back to *meh2- because of their cognates in Greek
35 In Jasanoff’s terminology, see Jasanoff 2003: 91-127.
36 With a regular loss of laryngeal before ♦i, cf. factitives in ♦-eh^ieIo- > Hittite -a(t) or ♦speh^i-e^) > tspat.
37 Apud Feist 1939: 341; Forrer was followed by Pokorny 1951-59: 704; Kloekhorst 2008: 541.
38 Latin maturus was first connected to Hittite mat- by L. Zgusta (1951: 452).
(eyxsoi)^ropo^, Old High German (Volk)mar (Germanic *merjaz, denominative *merjan ‘exalt’) and Slavic (Vladi)merb (unless borrowed from Germanic)39. But *mero- and *moro- can be easily reconciled with a reconstruction *meh2-, since these einzelsprachlich stems can be traced back to earlier *meh2ro- and *moh2ro-. Root ablaut in a thematic adjective is slightly troubling, but not unheard of; reconstructing *meh2ro— *meh2ro- ~ *moh2ro- seems to me to be a lesser sin than denying that *meh2- ‘to grow’ is related to *meh2- ‘big’.
My reason for maintaining both Yakubovich’s Hittite-Iranian connection and Forrer’s etymology of Hittite mai- (< *meh2-i-) has to do with the morphological set-up of Caland systems in the protolanguage: ‘much’, ‘big’ are typical Caland meanings and it is therefore plausible to assume that the root *meh2- ‘big’ had a Caland system. If this is the case, the verbal stem *meh2-i- associated with this root can be expected to have either a stative meaning ‘to be big’ or an inchoative one ‘to become big’ > ‘to grow’40. Slavic *matora- / *matera- ‘full grown’ and Latin maturus ‘full grown’ make it quite plausible that the meaning ‘grow’ was present already in the protolanguage41. As J. Jasanoff has shown42, *h2e-conjugation i-presents with 3 sg. in *-i-e are usually thematized to *-ie-ti in the daughter languages: a Hittite i-present in stative-inchoative function would therefore present a welcome parallel to the class of Indo-Iranian stative-inchoatives in -ya- whose firm place in the Caland system has recently been established by J. Rau43.
4. We can now return to a^aupoq ‘weak’. It seems that we now have both morphologically and semantically plausible etymology:
39
Yakubovich did not mention Oscan mais with its reflex of *a, perhaps tacitly assuming that it goes back not to *meh2is, but to *magios ( = Latin maius), cf. Meiser 1986: 38.
40 In Hittite it was renewed as miess- ‘to grow, to be born’ with a productive suffix.
41 But ‘to grow’ ^ ‘to become big’ (inchoative) is not the only meaning of Hittite mai: the meanings ‘to prosper, to thrive’ are best derived from the stative meaning ‘to be big, full grown’.
42 See Jasanoff 2003: 91-127.
43
See Rau 2009: 140-41. As J. Rau (p.c.) reminds me, Hittite mai would be the second example of an i-present in stative-inchoative function derived from a “Caland” root, the other one being *speh2- (Hittite ispai, Vedic (AV) sphayate, sphira- ‘fat’).
*meh2-i- 1) ‘to become big’, ‘to grow’; 2) ‘to be big’
^ verbal noun *meh2ur / -n- 1) ‘growth’; 2) ‘magnitude’
^ *n-meh2ur-o- ‘having no growth/magnitude’44 >
apaupo^ ‘weak’.
References
Николаев 2010 - Николаев А. С. Исследования по праиндоевропейской именной морфологии. СПб.: Наука, 2010.
Петлева 1973 - Петлева И. П. Праславянский слой лексики сербохорватского языка II // Этимология. 1971 (1973). С. 20-57.
ЭССЯ - Этимологический словарь славянских языков / Отв. ред. О. Н. Трубачев. М.: Наука, 1974-.
Acharyan 1971-1979 - Acharyan Hr. Hayeren armatakan bararan. Erevan: Erevani Hamalsarani Hratarakchutyun, 1971-1979.
Barrett 1964 - Barrett W. S. Euripides. Hippolytos. Oxford: Oxford Clarendon Press, 1964.
Bjorvand-Lindeman 2007 - Bjorvand H., Lindeman F. O. Vare arveord: etymologisk ordbok / 2nd ed. Oslo: Instituttet for sammenlignende kulturforskning, 2007.
Bucherer 1904 - Bucherer F. Anthologie aus den griechischen Lyriken. Gotha, 1904.
Cheung 2007 - Cheung J. Etymological Dictionary of the Iranian Verb. Leiden: Brill, 2007.
Clay 1993 - Clay J. S. Sappho 55 (Voigt): Going, Going, Gone // Electronic
v Antiquity. J993. P. 1.
Cop 1956 - Cop B. Beitrage zur indogermanischen Wortforschung I // Zeitschrift fur vergl. Sprachforschung. 1956. Bd. 74. S. 225-232.
Cronert 1912-1914 - Cronert W. Passow’s Worterbuch der griechischen Sprache, vollig neu bearbeitet. Gottingen: Vandenhoeck und Ruprecht, 1912-1914.
Derksen 1996 - Derksen R. Metatony in Baltic. Amsterdam; Atlanta: Rodopi, 1996.
DGE - Diccionario Griego-Espanol / Redactado bajo la direccion de F. Rodriguez Adrados. Madrid, 1980-.
Dunbar 1998 - Dunbar N. Aristophanes. Birds. Oxford: Oxford University Press, 1998.
Eichner 1973 - Eichner H. Die Etymologie von heth. mehur // MSS. 1973. Bd. 31. S. 53-107.
Falk-Torp 1960 - Falk H. S., Torp A. Norwegisch-danisches etymo-logisches Worterbuch, mit Literaturnachweisen strittiger Etymologien sowie deutschem und altnordischem Worterverzeichnis / 2nd ed. Oslo: Universitetsforlaget, 1960.
44 As usual, it is next to impossible to be sure whether dpaupoc; goes back to *n-meh2ur-o-, derived from *meh2ur, or to *n-meh2ur-o-, derived from a u-stem *mehu- (for a similar case see Nussbaum 1998). Note in any event that *-uro- may have been taken over from the antonym *peh2uro- ‘small’ (naupoc;).
Fick 1874-1876 - Fick A. Vergleichendes Worterbuch der indoger-manischen Sprachen. 2. Auflage. Gottingen: Vandenhoeck & Ruprecht, 1874-1876.
Fraenkel 1950 - Fraenkel E. Aeschylus. Agamemnon. Oxford: Clarendon Press, 1950.
Friis Johansen-Whittle 1980 - Friis Johansen H., Whittle E. Aeschylus: The Suppliants. Copenhagen: I Kommission hos Gyldendalske Boghandel, 1980.
Fortson 2007 - Fortson B. W. IV. The origin of the Latin future active participle // Verba Docenti. Studies in Historical and Indo-European Linguistics Presented to Jay H. Jasanoff by Students, Colleagues, and Friends / Ed. by A. J. Nussbaum. Ann Arbor; New York: Beech Stave Press, 2007. P. 83-96.
Handschur 1970 - Handschur E. Die Farb- und Glanzworter bei Homer und Hesiod, in den homerischen Hymnen und den Fragmenten des epischen Kyklos. Wien : Notring der wissenschaftlichen Verbande Osterreichs.
Heidermanns 1993 - Heidermanns F. Etymologisches Worterbuch der germanischen Primaradjektive. Berlin: de Gruyter, 1993.
Jasanoff 2003 - Jasanoff J. H. Hittite and Indo-European Verb. Oxford: Oxford University Press, 2003.
Jasanoff 2008 - Jasanoff J. H. The accentual type *vedo, *vedeti and the origin of mobility in the Balto-Slavic verb // Baltistica. 43. 2008. P. 339-379.
Kloekhorst 2008 - Kloekhorst A. Etymological Dictionary of the Hittite Inherited Lexicon. Leiden: Brill, 2008.
Kretschmer 1892 - Kretschmer P. Indogermanische accent- und lautstudien // Zeitschrift fur vergleichende Sprachforschung auf dem Gebiete der indogermanischen Sprachen. 1892. 31. S. 325-472.
Lundahl 1931 - Lundahl I. Ordet mo och dess etymologi // Goteborgs Hogskolas Arsskrift. 1930(1931). T. 36/3 ( = Minnesskirft utgiven av Filologiska Samfundet i Goteborg pa trettioarsdagen av dess stiftande den 22 oktober 1930). P. 20-37.
Luppino 1967 - Luppino A. Una formula omerica in Saffo // La parola del passato. 1967. T. 22. P. 286-291.
Martirosyan 2010 - Martirosyan Hr. K. Etymological Dictionary of the Armenian Inherited Lexicon. Leiden: Brill, 2010.
Matasovic 2009 - Matasovic R. Etymological Dictionary of Proto-Celtic. Leiden: Brill, 2009.
McKinlay 1957 - McKinlay A. P. On the way scholars interpret dpaupoc; // L’Antiquite classique. 1957. T. 26. P. 12-39.
Meiser 1986 - Meiser G. Lautgeschichte der umbrischen Sprache. Innsbruck, 1986.
Nagy 2000 - Nagy G. “Dream of a shade”: refractions of epic vision in Pindar’s “Pythian 8” and Aeschylus’ “Seven against Thebes” // Harvard Studies in Classical Philology. 2000. Vol. 100. P. 97-118.
Nauck 1872 - Nauck A. Kritische Bemerkungen VI // Bulletin de l’Academie imperiale des sciences de St.Petersbourg. 1872. T. 17. P. 180-275 ( = Melanges greco-romaines. 1974. T. 3. P. 207-344).
Nikolaev 2009 - Nikolaev A. S. The Germanic word for ‘sword’ and delocatival derivation in Proto-Indo-European II Journal of Indo-European Studies. 2009. Vol. 37. P. 4б1-488.
Noreen 1894 - Noreen A. Abriss der urgermanischen Lautlehre. Strassburg: Trubner, 1894.
Nussbaum 1998 - Nussbaum A. J. Severe problems II Mir Curad I Ed. by J. Jasanoff et al. Innsbruck. P. 521-538.
Page 1955 - Page D. L. Sappho and Alcaeus: An Introduction to the Study of Ancient Lesbian Poetry. Oxford: Oxford University Press, 1955.
Persson 1892 - Persson P. Om betydelsen och harledningen af det gr. араирос; (араирою), paupoc (раирою) jamte en exkurs om den grekiska, resp. indoeuropeiska u-epenthesen II Uppsalastudier tillegnade Sophus Bugge pa hans б0-ага fodelsedag den 5 januari 1893. Uppsala: Almqvist & Wiksells, 1892. P. 180-193.
Petersson 19^a - Petersson H. Baltisches und Slavisches. Lund; Leipzig: H. Ohlsson, 191б.
Petersson ^^b - Petersson H. Beitrage zur armenischen Wortkunde II Zeitschrift fur vergleichende Sprachforschung auf dem Gebiete der indogermanischen Sprachen. 191б. Bd. 47. S. 240-291.
Prellwitz 1905 - Prellwitz W. Etymologisches Worterbuch der griechischen Sprache. 2. Aufl. Gottingen: Vandenhoeck und Ruprecht, 1905.
Rau 2009 - Rau J. Indo-European Nominal Morphology: The Decads and the Caland System. Innsbruck, 2009.
Stokes 1901 - Stokes W. Irish etymologies II Zeitschrift fur celtische Philologie. 1901. Vol. 3. P. 4б7-473.
Stromberg 1944 - Stromberg R. Griechische Wortstudien: Untersuchungen zur Benennung von Tieren, Pflanzen, Korperteilen und Krankheiten. Goteborg: Elanders boktryckeri aktiebolag, 1944.
Troxler 19б4 - Troxler H. Sprache und Wortschatz Hesiods. Zurich: Juris Verlag, 19б4.
Tzamali 199б - Tzamali E. Syntax und Stil bei Sappho. Dettelbach: J.H. Roll, 199б.
Vine 2002 - Vine B. On full-grade ♦-ro- formations in Greek and Indo-European II Indo-European Perspectives I Ed. by M. R. V. Southern. Washington, D.C.: Institute for the Study of Man, 2002. P. 329-350.
de Vries 19б2 - Vries J. de. Altnordisches etymologisches Worterbuch. 2nd ed. Leiden: E. J. Brill, 19б2.
Weiss 1994 - Weiss M. On the non-verbal origin of the Greek verb v^siv ‘to be sober’ II Historische Sprachforschung. 1994. Vol. 107. P. 91-98.
Wilamowitz-Moellendorff 1895 - Wilamowitz-Moellendorff U. von. Euripides. Herakles. Bd. 2. Berlin: Weidmann, 1895.
Wilamowitz-Moellendorff 1913 - Wilamowitz-Moellendorff U. von. Sappho und Simonides. Berlin: Weidmann, 1913.
Yakubovich 2010 - Yakubovich I. S. Indo-European ♦ma- ‘to grow’ II Индоевропейское языкознание и классическая филология 14 I Отв. ред. Н. Н. Казанский. Т. 2. СПб.: Наука, 2010. P. 478-492.
Zupitza 1899 - Zupitza E. Etymologien II Beitrage zur Kunde der indogermanischen Sprachen. 1899. Bd. 25. S. 89-105.
