Greek Mythology and Poetics

  Use the following persistent identifier:

Chapter 7. Thunder and the Birth of Humankind

For surely you are not from an oak, as in the old stories, or from a rock.

To begin at the very beginnings is to begin with the oak and the rock, at least in the logic of the proverb, and it is for this reason, as we shall see, that the persona of Hesiod, reproaching himself for lingering too long at the beginning of beginnings in the Theogony, finally declares, with impatience:

ἀλλὰ τίη μοι ταῦτα περὶ δρῦν ἢ περὶ πέτρην;

Hesiod Theogony 35

But why do I have these things about the oak or about the rock?

In the interest of making our own beginning, let us without further delay proceed to the Baltic and the Slavic evidence. {182|183}

The derivation of tarḫu– from verb tarḫ– follows a familiar Hittite pattern: we may compare parku– ‘elevated’ from verb park– ‘lift’, ḫuišu– ‘alive’ from verb ḫuiš– ‘live’, and so on. [61] As for the derivation of Tarḫuna– from Tarḫu-, we may compare the form peruna– ‘rock’, derived from peru-’rock’; [62] both forms are used with the prefixed Sumerogram NA4, which also designates ‘rock’ or ‘stone’. The declension of Hittite peru– ‘rock’, {189|190} as we see from dative/locative peruni, reveals an r/n-stem added to the u-stem. [63] Hittite peru– ‘rock’ is cognate with Indic párvan– ‘joint [e.g. of sacrificial animal], knot [in plant]’. [64] The Indic noun can be explained as a derivative of the verb-root *per- ‘go to the end point, go over to the other side, arrive at the other side’, [65] with special reference tο the successful piercing through or cutting through of the body ‘s joints in the context of sacrifice. [66] We may compare the Greek verb peírō, which can mean either ‘pierce’ when the object is the body of a victim or ‘cross over’ when the object is a body of water (as reflected in such derivative nouns as póros). [67] These meanings are pertinent to the semantics of the Hittite verb tarḫ– ‘conquer, overpower, overcome’, from which the name of Tarḫu– the Storm-God is derived: the Indo-European root *terh2– underlying this Hittite verb also carries the meaning ‘cross over’ as in the Indic aptúr– ‘crossing over the water’, [68] and this meaning applies also in specific contexts of immortalization, as in the Greek néktar. [69] Given the parallelism between the sense of ‘cross over’ in the verb-root *terh2– and the sense of ‘go to the end point, go over to the other side, arrive at the other side’ in the verb-root *per-, we arrive at a better understanding of the etymology of Greek Ēlúsion, designating both the place where a thunderbolt has struck and the place of immortalization, ‘Elysium’: [70] this noun-formation is derived from –ḗlutho-, from the conjugation ἔρχομαι/ἐλεύσομαι in the sense of ‘arrive’. [71] Also, given the parallelism between the Hittite forms Tarḫu-/Tarḫuna-, designating the Storm-God, and peru-/peruna-, meaning ‘rock’, we may be ready to connect Hittite peruna– ‘rock’ with Slavic perunŭ ‘(god of) thunderbolt’, just as we have connected Latin quercus ‘oak’ with Baltic (Lithuanian) perkū́nas ‘(god of) thunderbolt’. As we have already seen, rocks as well as oaks were sacred to the god of the thunderbolt. [72]

As further illustration of the weapon/target ambiguity, we may consider the Indic noun áśman-, cognate of Greek ákmōn and Lithuanian akmuõ. This word is used in the RigVeda to designate the weapon of Indra (2.30.5, 4.22.1, 7.104.19). The activities of Indra, this national war-god, are as a rule described in a specialized language that is so highly stylized that it tends to blur the naturalistic aspects of his background as thunder-god, more visible in his counterpart Parjánya-. [83] Nevertheless, the basic naturalistic attributes of Indra persist in the RigVeda: he gives rain (4.26.2, etc.), lightning comes from him (2.13.7), he is likened to a thundering cloud-driver (6.44.12), and he is specifically described as equal to Parjanya at raintime (8.6.1). Indra’s weapon is predominantly called the vájra-, which, in keeping with the stylistic specialization of his descriptions, has become Indra’s personal distinguishing feature in the diction of the RigVeda. Like its owner, however, the vájra-, too, has natural attributes that persist: it thunders (1.100.13) and roars (2.11.10). With this vájra– of his, Indra conventionally strikes boulders and thereby releases water or light—a theme so common that any listing of the attestations would be superfluous. [84] Suffice it to note here that one of the words for ‘boulder’ in the RigVeda is áśman– (1.130.3, etc.), the same word that can also designate Indra’s weapon (again, 2.30.5, etc.). [85] {192|193}

The time has come to ask why oaks and rocks should be singled out for a sacral affinity with the thunderbolt to such an extent that their designations are interchangeable in various Indo-European languages. From our own secular standpoint, it is obvious that rocks, boulders, trees, hills, or mountains are targets of lightning by virtue of their elevation or prominence. But there are other factors as well in attracting the stroke of lightning. I would be out of my field in attempting a strictly scientific discussion about these other factors, but they can be in any case easily intuited even by the unscientific mind. For example, it is an observable fact that different kinds of trees have significantly different degrees of susceptibility to being struck by lightning. In a quaint experiment conducted by the Lippe-Detmold’sche Forstverwaltung over the years 1879 to 1890, the following statistics emerge for the susceptibility, {195|196} in a given forest, of certain species of tree to lightning strikes:

  % of all trees in forest number of lightening strikes
oak 11% 56
beech 70% 0
spruce 13% 3 or 4
fir 6% 20 or 21

As we shall now see, the infusion of thunder-fire into wood and stone is a sexual and anthropogonic theme. Let us begin with instances of overtly creative themes associated with the thunderbolt. In Indic {196|197} mythology, for example, Indra’s vájra-, his stylized thunderbolt, is not only destructive but also procreative. [111] The Iranian cognate, vazra-, forms a derivative vazraka-, which means ‘endowed with generative power’ in Old Persian. [112] The radical *u̯eg- of vájra/vazra– recurs in the Rig-Vedic word vā́ja– ‘generative power residing in vegetation, cattle, etc.’. [113] It also recurs in Latin uigeō ‘thrive’ and uegeō ‘quicken, arouse’. [114] We may note in this connection the Old Norse lore about Thor’s Hammer, which hallows the laps of brides and has the power of bringing his dead goats back to life. [115] Or we may note a Lithuanian belief, reported by the cleric Matthäus Praetorius (late seventeenth century), that lightning could beget human children in the vicinity where it strikes. [116] Or again, if lightning strikes in the daytime as a child is being born, he will thrive; if it strikes at night, the child will die. [117] If a man is struck down in a thunderstorm that is heading west, he dies as a favorite of God; if the thunderstorm was heading east, he has died on account of his sins. [118] Within the framework of this presentation, however, I cannot do justice to the vast subject of the thunderbolt’s destructive/creative ambivalence in Indo-European lore. [119] My main purpose instead is to explore specifically the Indo-European traditions concerning the association of the thunderbolt with trees and rocks, and how the action of a thunderstroke on these materials was believed to be sacral and, more than that, creative.

A belief in the creative and even anthropogonic powers of trees or rocks is indirectly attested in a Greek proverb: surely you or I or anyone else today, the saying has it, were not created either from oaks or from rocks (e.g. {197|198} Plato Apology 34d, Republic 544d; Plutarch Consolation to his Wife 608c; Philostratus Images 2.3.1). By implication, as we shall see, earlier humans had just such origins.

In Hesiodic poetry as well, there is a fastidious attitude toward treating the theme of oaks and rocks with any references that would go beyond mere allusion. There is a passage in the Theogony (31-34) where the poet has just told how the Muses infused in him the power to sing about {198|199} past and future things, and about the origins of the gods; also, to start and end the song by singing of the Muses themselves. The poet then breaks off with these words:

ἀλλὰ τίη μοι ταῦτα περὶ δρῦv ἢ περὶ πέτρην;

Hesiod Theogony 35

But why do I have these things about the oak or about the rock?

With this utterance, the narration is pausing to take a self-conscious look at the point that has been reached so far in the composition of the Theogony. In the next verse, the break is followed up with Μουσάων ἀρχώμεθα ‘let us start with the Muses’, the same expression that had inaugurated the Theogony at verse 1. Thus the narration has come full circle from Theogony 1 to 36, and Hesiod “has to make a fresh start on the same lines as before.” [
121] Verse 35 actually anticipates that Hesiod is about to make this fresh start with verses 36 and following. For Hesiod to ask in verse 35 why he has “these things about [= going around] [122] the oak or about the rock” is the equivalent of asking why he has lingered at the beginning of beginnings. “Why am I still going around, as it were, the proverbial oak or rock? Let me proceed at last by starting out again!”

Instead of going further with illustrations, however, I simply return to the cenital point lo be made. We have observed an ambivalence in the application of Indo-European *per(kw)u- to oaks in some languages {200|201} (Latin quercus) and to rocks in others (Hittite peru). This etymological ambivalence, we now see, is matched by the thematic ambivalence in the obsolescent ancient Greek proverb alluding to the myth that humankind originated either from oaks or from rocks—whichever of the two. The unifying theme that resolves these two cases of ambivalence, I am suggesting, is the creative action of the Indo-European thunder-god, whose very name is formed from the derivatives of *per(kw)u- in Baltic (Perkū́nas) and Slavic (Perunǔ). {201|202}


[ back ] 1. Frazer 1930.224-225; also 90, 92, 151, 155 (trees) and 106, 131, 187-188 (rocks/stones).

[ back ] 2. Frazer p. 90.

[ back ] 3. Cf. Frazer pp. 220-221.

[ back ] 4. Bachelard 1949, esp. pp. 45-47.

[ back ] 5. Cf. Dworak 1938, esp. p. 1.

[ back ] 6. For the difficult task of establishing the etymological links between these words, I cite the parthfinding work of Ivanov 1958 and Ivanov and Toporov 1970, folloing Jakobson 1950, 1955; cf. Watkins 1966.33-34, 1970.350, 1974.107. Book-length treatment in Ivanov and Toporov 1974.

[ back ] 7. Cf. REW s.v. perún.

[ back ] 8. The testimonia have beeen assembled by Gimbutas 1967.741-742.

[ back ] 9. Cf. Darkevič 1961.91-102.

[ back ] 10. Cf. Ivanov and Toporov 1968, 1970.

[ back ] 11. Ivanov and Toporov 1970.1182.

[ back ] 12. Seržputovski 1930.26 (I no. 268); cf. Ivanov and Toporov p. 1194.

[ back ] 13. Cf. the expression Pērkōns męt savu milnu ‘Pērkōns throws his mace’ (Ivanov and Toporov p. 1195). Note that milna ‘mace’ is related to Old Norse mjǫllnir, the word for the hammer of Thor the thunder-god. Note, too, that Thor’s mother is Fjǫrgyn (from *perkunī). On these forms: Ivanov 1958.104.

[ back ] 14. Balys 1937.163 nos. 233-236; also p. 149 nos. 4-5. Cf. Ivanov Toporov p. 1182.

[ back ] 15. For an attestation of this believ, see the first-person account of the cleric Matthäus Praetorius (late seventeenth century) of his encounter with some Lithuanian woodcutters, as reprinted in Mannhardt 1936.533-535. For the relative reliability of Praetorius, see Mannhardt pp.519-520.

[ back ] 16. Balys 1937.163 no. 246; cf. Ivanov and Toporov p. 1184.

[ back ] 17. Balys p. 163 no. 241; cf. Ivanov and Toporov p. 1184

[ back ] 18. Šmits 1940.1401; cf. Ivanov and Toporov p. 1184

[ back ] 19. Balys 1937.158 no. 141 and p. 197 no. 802; cf. Ivanov and Toporov pp. 1193-1194.

[ back ] 20. For a detailed account by Matthäus Praetorius, see Mannhardt 1936.539-540 (cf. Ivanov and Toporov p. 1189). Perkūnas was venerated with perpetual fires fueled by ‘oak wood (for documentation, see Mannhardt pp. 196, 335, 435, 535, etc.). When Christian zealots extinguished such perpetual fires, the natives believed that Perkūnas would freeze (Mannhardt p. 436).

[ back ] 21. Mannhardt p. 196; cf. Ivanov and Toporov p. 1187. For a balanced account of Grunau’s basic reliability, see Krollmann 1927.14-17. Without reading Krollmann, we would be prone to overinterpret the severe judgment of Jaskiewicz 1952.92-93, who was primarily concerned with the unreliability of a later writer, Jan Lasicki (late sixteenth century). Krollmann argues cogently the unlikelihood of Grunau’s having “invented” a Prussian system of gods modeled on a Nordic scheme. It strains credulity to imagine that this wandering beggar-monk (who even spoke Prussian himself) would have created a pastiche based on Adam of Bremen (Krollmann pp. 15-17). Krollmann’s inference, however, that the Prussian religious practices described by Grunau were borrowed from Nordic culture (p. 17) is gratuitous, As for Jaskiewicz, I find in his brief discussion of Grunau no facts to support his contention that Grunau’s account of Perkūnas and the ‘oak is mere phantasmagoria (pp. 92-93). For a balanced evaluation of Grunau’s reliability, see Puhvel 1974.

[ back ] 22. Ivanov and Toporov pp. 1183-1184.

[ back ] 23. Seržputovski 1930.9 no. 49; cf. also p. 8 nos. 37-48. Cf. also Ivanov and Toporov pp.1193-1194.

[ back ] 24. Balys 137.159 nos. 155-164; cf. Ivanov and Toporov p. 1194.

[ back ] 25. Mannhardt 1936.539-540.

[ back ] 26. Ivanov and Toporov p. 1193.

[ back ] 27. REW s.v. perún.

[ back ] 28. Also Russian peru/prat’, Czech peru/prati, etc.; cf. Lithuanian per̃ti ‘strike’. For a typical example of the agent-suffix –unŭ, cf. Russian begún ‘runner’ from begat’ ‘run’.

[ back ] 29. Schulze 1929.287. Cf. also Latin tribūnus from tribus, lacūna from lacus, and so on.

[ back ] 30. DELL, s.v. For a survey of Germanic cognates meaning ‘oak’ or ‘fir’, see the discussion of Friedrich 1970.136-137. For the semantic shift from ‘oak’ to ‘fir’, see Friedrich p. 13Gn30; also Vendryes 1927.314-315 and Güntert 1914.214.

[ back ] 31. Specht 1932.215: e.g. karaliū́nas ‘prince’ from karãlius ‘king’.

[ back ] 32. Otrębski 1965.206; the suffix –ūnas too may be denominative (Otrębski p. 207), but the point remains that –ū́nas cannot be deverbative.

[ back ] 33. LEW s.v.

[ back ] 34. LEW s.v.; cf. Meillet 1926.171.

[ back ] 35. Meillet p. 171 and LEW s.v. For the derivative orot ‘thunder’, see Lidén 1906.88ff.

[ back ] 36. Meillet p. 171 and KEWA s.v.

[ back ] 37. We may note that –áni– is a deverbative formant of abstract nouns and adjectives, suchas śaráṇi– ‘malice’ from śr̥– ‘destroy’ and sakṣáni– ‘overpowering’ from sah– ‘overpower’.For more examples, see Wackernagel and Debrunner 1954.207, where it is pointed out that abstract nouns in *eni- arc also a productive category in Germanic. (The type –áni– becomes infinitival in Indic, and some nouns in –ani– seem to be formed by way of infinitives in –áni-: see Renou 1937, esp. pp. 73-78.) There are two especially interesting examples: áni– ‘thunderbolt’ and aráṇi– ‘wood for obtaining fire by friction’ (on which see p. 156). In neither case has the respective verb of these derivatives even survived in Indic. Sometimes the suffix –áni– is found with the accentuation –aní-, as in kṣipaní– ‘stroke of the whip’ from kṣip– ‘hurl’ and dyotaní– ‘brightness’ from dyut– ‘shine’. The point is, the suffixes –anyà– (= –anía-) and –ánya– seem to be derivatives of –aní– and –áni– respectively. (For the derivation of i̯o-stems from i-stems, cf. Wackernagel and Debrunner pp. 778, 804, 816-817; also Benveniste 1935.73-74.) For examples of the suffix –anya-, we may adduce nabhanya– ‘bursting forth’ from nabh– ‘burst’ and abhyavanyà– ‘apportioning’ from abhyavadhā– ‘apportion’; cf. Wackernagel and Debrunner p. 212. In sum, I interpret parjánya– as ‘striking, the striker’, from *perg- ‘strike’.

[ back ] 38. Jakobson 1955.

[ back ] 39. Jakobson 1955. LEW 574 also connects the following words containing the root *per-: Lithuanian pérgas ‘Einbaum, Fischerkahn’, Old English fercal ‘bolt’, Latin реrgula ‘projection from an edifice’. We may perhaps add Czech prkno ‘board’. For the semantics, cf. Trautmann 1906. On Indic parkatī, see KEWA s.v.

[ back ] 40. Here we have the central point of Ivanov and Toporov 1970.

[ back ] 41. Meid 1957.126; Ivanov 1958.105; Güntert 1914.213.

[ back ] 42. Ivanov 1958.104-104, 107; for a discussion of related forms in Germanic, see Feist 1939 s.vv. fairguni and fairhvus; I draw attention to the delabialization of *kw to *k: *perkwun- to *perkun-.

[ back ] 43. Ivanov 1958.104-105, 107; Watkins 1966.33-34; Meid 195.284-285.

[ back ] 44. Ivanov 1958.104.

[ back ] 45. Cf. p. 183n13. The formal type *perkwŭ-ni̯o- is worth contrasting with the type *perkwū-no- as in Lithuanian perkū́nas, where the derivation from *perkwu- is accompanied by lengthening of the *-u-. The same kind of lengthening, as we have already noted, occurs in Latin: Роrtūnus from portus, tribūnus from tribus, and so on. A comparable kind of non-lengthening pattern as in *perkwŭ-ni̯o- is evident in Indic: árjuna– ‘bright’ (cf. Greek árgǔros, árgǔphos); cf. also the inner-lndic derivative patterns śmaśrǔ– ‘bearded’ from śmaśrǔ-’beard’, dārúna-’sturdy’ from dā́ru– ‘wood’, etc. (cf. Wackernagel and Debrunner 1954.485-486, 734; Meid 1956.270, 280).

[ back ] 46. Jakobson 1955; cf. also Ivanov 1958.107-108, pace Vaillant 1948.

[ back ] 47. Jakobson 1955.

[ back ] 48. Mansikka 1922.305; Jakobson 1955.616.

[ back ] 49. See p. 184.

[ back ] 50. Ivanov 1958.107.

[ back ] 51. Mansikka 1922.65, 380; Ivanov p. 107.

[ back ] 52. Jakobson 1955.615, Gimbutas1967.742.

[ back ] 53. Ivanov 1958.106-107. We may compare the variation in the Slavic word for ‘wormwood’, *pelynŭ/pelunŭ, as attested in Old Church Slavonic pelynŭ, Polish piołyn/piołun, Czech pelyn/pelun, and so on. Cf. Būga 1959 [1921].332; cf. also Meid 1956.273-274.

[ back ] 54. I reconstruct Slavic –ynŭ/-unŭ as *-ūnos/*-ōunos, on the basis of comparative evidence from Baltic, especially Latvian. For the clearest example available, I note that the Latvian u-stem noun vìrsus ‘summit’ has the variant derivatives vìrsūne and vìrsuone, both likewise meaning ‘summit’ (LDW 4:616). We may compare Lithuanian viršū́ne ‘summit’, from viršùs ‘summit’. Such an alternation –ūn-/-uon– clearly suggests an earlier *-ūn-/-ōun- (Endzelin 1923.235, 240; cf. Meid 1956.276 on the type υἱύς/υἱωνός; the negative arguments of Schmeja 1963.40-41 are based mainly on the relative dearth of positive evidence in Greek). Latvian also shows a third variant, vìrsaune (LDW 4:610-611), which is significant because inherited *ōu has a bivalent reflex in Baltic, au as well as uo (Stang 1966.47-48, 75-76). We see the ū/uo/au alternation not only in derivatives of u-stems such as vìrsūne/vìrsuone/vìrsaune but also in the actual declension of both Lithuanian and Latvian u-stems (Slang pp. 75-76). And it so happens that we find the same ū/uo/au alternation in the attested Latvian variants of the word for ‘thunderbolt’, pḕrkūns/pḕrkuons/pḕrkauns (LDW 3:208-209); besides these o-stems, we also find the i̯о-stems pę̄ rkūnis/pę̄rkuonis/pę̄rkaunis (LDW 3:208-209).

[ back ] 55. See the previous note. Of these three formations, it is pḕrkuons that prevails in the standard language (spelled pērkōns), largely because agent-nouns in –uons are a productive category in Latvian (the suffix –uonis is likewise productive in Latvian; cf. Lithuanian –uonis, which is produtive, whereas –uonas, cognate of Latvian –uons, is not). Specht I932 (240-241, 259, 264-265, etc.) has demonstrated that the suffix –uons of such agent-nouns is derived ultimately from *-ōn- (as in Greek -ων); the *-ōn- is clearly attested in older Lithuanian –; the replacement pattern that prevailed in this language is not the o-stem –uonas but the i̯o-stem –uonis (unlike Latvian, where –uons and –uonis coexist). Specht’s demonstration, however, need not lead to the inference that Latvian pḕrkuons, as distinct from pḕrkūns, is not an inherited form. It is simply a matter of phonological ambiguity in Baltic (both *ōu and *ō yield uo), which has led to a morphological reinterpretation. The variant pḕrkuons prevails over pḕrkūns because its suffix –uons, which I reconstruct as *-ōunos, now has the same familiar shape as the productive formant of agent-nouns, –uons as derived from *-ōnos. Let us contrast a situation where we find no such inherent parallelism, as in the case of vìrsūne/vìrsuone/virsaune, discussed in the previous note: bere the form that prevails is vìrsaune, not vìrsuone (for all the forms, see LOW 4:610-611). In considering the displacement of pḕrkūns by pḕrkuons in Latvian, we should note, too, that this language does not have a productive category of agent-nouns in –ūns. Conversely. Lithuanian perkū́nas can prevail over perkuonas because the language has no productive category of agent-nouns in –uonas (from *-ōnos); rather, the productive suffixes are –uonis and –ūnas. For vestiges of perkuonas (and viršuone) in Old Lithuanian, see Specht p. 265.

[ back ] 56. Just as Latvian pḕrkuons prevails over pḕrkūns because of the productive agent-suffix –uons (see previous note), so also we may say that Slavic perunŭ prevails over *perynŭ because its suffix –unŭ, which I reconstruct as *-ōunos, is shaped like an agent-suffix that is productive in Slavic, –unŭ. In this case, I prefer to reconstruct the deverbative agent-suffix also as *-ōunos, the same suffix that is derived from u-stem nouns. In other words, deverbative –unŭ was once denominative (cf. Specht 1932.268). We may compare Lithuanian –ūnas and –uonis, which are still denominative as well as deverbative, as distinct from –ū́nas, which is formally marked for an exclusively denominative category (cf. Specht pp. 240-241; also Meid 1956.268-270).

[ back ] 57. This is not to say that e.g. *terh2u- cannot be a verb-formation (cf. Hittite tarḫuzzi, Indic tárute, tū́rvati, etc.).

[ back ] 58. Laroche 1958.90. Also Watkins 1990.

[ back ] 59. Laroche pp. 91-95; on Tarḫunt– and the thunderbolt, see Laroche p. 95. Cf. Watkins 1974.107. Cf. also p. 131.

[ back ] 60. Laroche pp. 93, 94.

[ back ] 61. Laroche p. 90. As for the vocalism of tarḫ-, etc., Kuryłowicz 1958.228 observes that roots ending in –erC tend to generalize the zero-grade in Hittite (-R̥C> –arC, where R = r, l, m, n and C = other consant).

[ back ] 62. Laroche p. 90. Also spelled piruna– and piru-.

[ back ] 63. Hoffmann 1975 [1974].332,336.

[ back ] 64. Hoffmann pp. 332, 336.

[ back ] 65. Bergren 1975.62-101, esp. p. 95.

[ back ] 66. Bergren pp. 67-78, esp. pp. 68-69 on RigVeda 1.61.12, where Indra hurls his thunderbold at Vr̥tra in order to sever his joints (noun párvan), much as the joints of an ox are severed; cf. also Hoffmann 1975 [1974].332.

[ back ] 67. Cf. DELG 871. Detailed discussion of the semantics in Bergren pp. 95-101.

[ back ] 68. See p. 139.

[ back ] 69. See p. 139. Cf. aso. p. 156n49 on Indic taráni– in the sense of ‘ship’.

[ back ] 70. See p. 140.

[ back ] 71. On the level of form, this derivation of Ēlúsion is corroborated by DELG 411; my interpretation of the meaning, however, differs from that of DELG.

[ back ] 72. See pp. 184ff.

[ back ] 73. Cf. also the derivative naktìgonis ‘night-herder of horses’: LEW 152.

[ back ] 74. For the Ullikummi texts, see Güterbock 1952, esp. pp. 37, 146-147.

[ back ] 75. Ivanov 1958.110, Ivanov and Toporov 1970.1196-1197.

[ back ] 76. Grassmann 1873 s.v. han-, the attestations labeled “1.”

[ back ] 77. Jakobson 1955.616, with a list of tabu-variants.

[ back ] 78. Jakobson 1950.1026.

[ back ] 79. For another example of Hittite instrument-noun in –uzzi-: išḫuzzi– ‘belt’, vs verb išḫiya– ‘bind’.

[ back ] 80. Sec Güterbock 1952.6. The Storm-God is written with the Sumerogram dU, but the Ullikummi texts (as well as others) reveal the Hittite ending –unaš; сf. Güterbock p. 4n14. The full form is probably Tarḫunaš (Laroche 1958.94-95).

[ back ] 81. Whitman 1970, esp. pp. 39-40.

[ back ] 82. Rostowski’s history of the Jesuit Order in Lithuania (1538): see Mannhardt 1936.435 and Reichelt 1913.26. On the related subject of Iuppiter Lapis, see Schwenck 1859.393-394. Cf. also the Laconian cult stone known as Zeùs Kappṓtās (Pausanias 3.22.1).

[ back ] 83. On this figure, see p. 180. For a balanced discussion of the naturalism surrounding the Parjánya– figure, see Lommel 1939.38—44. For a contrast of the naturalistic/nonnaturalistic descriptions of Parjanya/lndra, see Oldenberg 1917.137. The relative dearth of naturalism in descriptions of Indra results from intense stylistic elaboration and evolution, not shared by the far less developed Parjanya figure.

[ back ] 84. For a survey, see Rechelt 1913.34-37.

[ back ] 85. In RigVeda 2.12.3, Indra brings forth fire áśmanor antár ‘between two rocks’. Ivanov and Toporov 1970.1195-119 compare the Byelorussian theme where Perun (piarun) rubs two gigantic millstones together, thus producing thunder and lightning.

[ back ] 86. See pp. 189ff.

[ back ] 87. See again Lommel 1939.42-43. In a few passages, áśman– allows the interpretation‘sky’ (e.g. RigVeda 7.88.2; but see Geldner 1951 2:259 and Kniper 1964.111n80). Reichelt1913 has argued that áśman– could be used with the meaning ‘sky’ as well as ‘stone’ becausethe sky was thought to have a stone vault once upon a time. We may compare especiallyAvestan asan-/asman’rack, stone, sky’ (Bartholomae 1904.207-208). For typological parallels from Africa, see Baumann 1936.146-147. Although Reichelt’s presentation is persuasive, there is another way to explain the meaning ‘sky’. If ‘rock’ equals ‘cloud’ by way of metaphor, the notion of ‘cloud’ could easily evolve into ‘sky’. We may compare Hittite nepiš ‘sky’ and Slavic nebo ‘sky’ vs. Indic nábhas ‘cloud’ and Greek néphos ‘cloud’; cf. alsoMiddle English sky ‘cloud’ or ‘sky’.

[ back ] 88. For Slavic parallels: Ivanov and Toporov 1970.1193. For typological parallels in African myth and ritual: Hocart 1936.56.

[ back ] 89. Lommel 1939.42-43.

[ back ] 90. See Geldner 1951 1:36, etc.

[ back ] 91. See Benveniste and Renou 1934.147n1.

[ back ] 92. In the same stanza, Indra’s weapon is called áśman-: see Reichelt 1913.44-45. Cf. also RigVeda 2.30.5, 4.22.1. On the Germanie theme οf the whetstone as a symbol of authority, see Mitchell 1985.

[ back ] 93. Benveniste and Renou 1934.138.

[ back ] 94. For Slavic parallels, see Ivanov and Toporov 1970.1193-1195.

[ back ] 95. Ivanov 1958.110 adduces the name of the Anatolian deity Pirwa– (on whom see Otten 1951), deriving it from *peru̯-o-. We may note too the Hittite collocation ḫékur irwa (cf. Goetze 1954.356n54), with the ḫékur preceded by the Sumerogramd, which designates ‘rock’ or ‘stone’, and will) irwa once preceded by the Sumerogramd; the noun ḫekur means ‘summit, mountain’.

[ back ] 96. Güntert 1914.215-216.

[ back ] 97. Güntert pp. 215-216, 221.

[ back ] 98. Güntert p. 216. Outside of keraunós, we find no instance of suffixal –aunos in Greek. The word púraunos (Pollux 6.88, 10.104) must be a compound formed with the verb aúō (on this word see Borthwick 1969).

[ back ] 99. Güntert p. 216.

[ back ] 100. DELG 519. From the evidence of such pairs as daíō (from *dau̯i̯ō) vs. daízō, we may expect an earlier form *kerau̯i̯ō: cf. GEW 822.

[ back ] 101. For reflexes of *h2u̯- as au in Greek roots, cf. both GEW and DELG s.v. kaíō (cf. also the discussion of Schmeja 1963.29-32).

[ back ] 102. See KEWA s.v. śr̥nā́ti.

[ back ] 103. Cf. artkópos ‘baker’ from *artokwopos, by metathesis from *artopokwos (on which see DELG 118)

[ back ] 104. Alternatively, terpikéraunos may be explained as the reflex of an expressive reduplication *kwerpi-ker(p)aunos, with the dissimilation of *kw…*kw to *kw…k, from an earlier form *kwer(p)aunos, which in turn would be a tabu-metathesis of *per(kw)aunos (cf. Watkins 1970.350).

[ back ] 105. Grimm 187 3:64. For a survey of connections between the oak and Thor (as well as other Germanic equivalents), see Wagler 1891 2:43-46.

[ back ] 106. See p.183.

[ back ] 107. Kuhn I886, with a discussion of the evidence collected by J. Grimm.

[ back ] 108. See pp. 192-193n85; Ivanov and Toporov 1970.1195-1196.

[ back ] 109. Ivanov and Toporov pp. 1195-1196.

[ back ] 110. In the case of wood, we have just noted a pattern of preference for oak wood as the sacral material for rubbing fire, a pattern that can be correlated with the observably stronger attraction of oak trees to lightning.

[ back ] 111. Survey of relevant passages in Gonda 1954.32-55, esp. pp. 36-37.

[ back ] 112. Liebert 1962.127, who also refutes the reinterpretation as vazarka– (e.g. Bartholomae 1904.1389-1390).

[ back ] 113. Gonda 1954.43ff; Liebert p. 145. Otherwise Watkins 1986.325 and 327, with the· references at n10.

[ back ] 114. Gonda pp. 43ff. For the vocalisms uig-/ueg-, see Watkins I973a.

[ back ] 115. Davidson 1965.11-14.

[ back ] 116. Reprinted in Mannhardt 1936.538.

[ back ] 117. Mannhardt p. 538.

[ back ] 118. Mannhardt p. 538.

[ back ] 119. More on this subject in Gonda 1954.36-37. Also, we may note that the derivation of neuter śárīra– ‘body’ from the root *kerh2– of śr̥nā́ti ‘shatter’ is accepted as plausible in KEWA s.v. śárīra– (on the associations of the verb śr̥nā́ti ‘shatter’ with the thunderbolt at RigVeda 3.30.17, see p. 195). We may compare the etymological explanation of Latin corpus as if derived, albeit indirectly, from the root *per-kw– (see Vendryes 1927.315). It may be possible to link the verb śr̥nā́ti not only with śárīra– ‘body’ but also with śáru– ‘missile, arrow’ (on which see p. 194), a word used for comparisons with the thunderbolt in the RigVeda (1.172.2, etc.). Perhaps also with śará– ‘reed’ and śáras– ‘ashes’ (cf. KEWA s.v.). In view of the latter meaning, ‘ashes’, it may be pertinent to cite the reflex of the root *perkw– in Lithuanian pir̃kšnys ‘glowing ashes’ (cf. LEW s.v. pir̃kšnys) and possibly also in Old Irish richis ‘glowing coals, live ember’.

[ back ] 120. See Odyssey ix 507, xiii 172. For the mantic connotations of thésphata, see especially ix 507, xi 151. In connection with the mantic concept, we may note the form of the place-name, Perkṓtē, as attested at Iliad II 835. A mántis ‘seer’ by the name of Mérops is identified as Perkṓsios ‘from Perkṓtē̕ at Iliad 831-832 (Μέροπος Περκωσίου, ὃς περὶ πάντων | ᾔδεε μαντοσύνας). The form Perkṓtē perhaps be reconstructed as *perkō[u]tā (for the phonology, cf. Vine 1982.42-43 on Greek plōtḗr), to be derived from an earlier u-stem noun *perku- (from *perkwu-) ‘oak’, We may note too that méropes is a common Homeric epithet for ánthrōpoi ‘humans’ (cf. Koller 1968) and that the sons of Mérops Perkṓsios are said to hold sway over a place called Pitúeia, at Iliad II 829. The latter name is surely derived from the noun pítus ‘pine’. We may perhaps compare the semantic oscillation between ‘oak’ and ‘fir’, on which see p. 185n30. What we see in these associations are perhaps traces of an ancient local myth that equated the First Man with the First Mántis ‘Seer’. On the anthropogonic theme of First Man as First Sacrificer, cf. pp. 110 and following. Given the anthropogonic themes inherent in the possible etymology of ánthrōpos as ‘he who has the looks of embers’ (p. 151n30), we may consider the possible etymology of mérops as ‘he who has glowing looks’; cf. Merópē as the name of a star at Hesiod F 169.3 MW. I suggest that Μαῖα, the name of another star mentioned in Hesiod F I69.3, be emended to Μαῖρα = Maîra, from root *mer- as in marmaírō ‘glow, flash’; cf. maī́rlē ‘embers of charcoal’. The proposed explanation of mérops as ‘he who has glowing looks’ may be pertinent to the discussion of Indic márya– at p. 250.

[ back ] 121. West 1966.170.

[ back ] 122. West p. 169 argues that περί + accusative regularly conveys a positional (‘around’) rather than a conceptual (‘about’) sense in early poetic diction. Perhaps such a formulation is too restrictive: the second sense can he a metaphorical derivative of the first.

[ back ] 123. What follows at Iliad XXII 127-128 is a description of sweet-talk between unmarried lovers, as if such a pair would take their conversation all the way back to the oak and the rock.

[ back ] 124. There is a multiude of African examples collected by Baumann 1936.224-235 (trees), 219-220 (rocks). For a survey οf European examples, cf. Vadé 1977. For traces in Semitic myth: Dirlmeier 1955.25-26. For a convenient bibliography on the general theme of Petra Genitrix, see Eliade 1962.208.

[ back ] 125. Specht 1944. Cf. also Lincold 1985.188n29, with emphasis on parallelisms between cosmogonic and sociogonic themes, especially in Germanic traditions.

[ back ] 126. See also Vendryes 1927.

[ back ] 127. Cf. Loth 1920.122.

[ back ] 128. Cf. Meulen 1907.55-72, 121-169. For a useful statement on tree animism, see esp. his p. 127. For a discussion of myths where humankind originates from the ash tree, see Shannon 1975.44—48, 57, 70. Shannon’s book shows convincingly that such myths are linked to the theme of Achilles’ ash spear in the Iliad. He also points out that, in the Hesiodic Work and Days, it is the gods in general who create the first and second generations of humankind (109-110, 127-128), but it is Zeus in particular who creates the third generation, and this third creation emerges specifically out of ash trees (143-145). Such an association of Zeus himself with the process of creation out of ash trees is significant in view of his epithets keraunós, terpikéraunos, and argikéraunos, as discussed at pp. 194-195.