Evidence for the meaning of the Indo-European Root *nes-

1. Introduction

The purpose of this chapter is to explore the background of the Greek root nes-. In Greek itself three factors are involved: the meaning of the root, the interpretation of this meaning, and the presence of the root in the word nóos. These factors have been considered at length, and the problem now is to determine what in each case was inherited by Greek from Indo-European.
One can approach this problem only by means of the comparative method. In this chapter, therefore, I shall be concerned with attestations of the root nes- in languages other than Greek. The evidence to be considered bears, first of all, on the meaning “return to life” of Greek nes-; the cardinal evidence for this is provided by Germanic languages, which will therefore be considered first. As for the involvement in sun symbolism of the root nes-, the main evidence is provided by Sanskrit, although the meaning “return to light” seems to have been preserved in Germanic as well, and also in Albanian. Attention will therefore be paid to this in section 2, “Germanic,” and a brief section 3 will then be devoted to Albanian. In section 4, “Indic,” the context of sun symbolism will be explored more fully.
The final factor is the derivation of Greek nóos from nes-. In the Indic section a case will be made that this root already implied “intelligence” in Indo-European. {125|126}

2. Germanic

The modern German word genesen, “get well, recover,” contains the Indo-European root nes-, and, by its meaning, still bears witness to the original meaning of the root. With only this word in mind one can understand what led Sigmund Feist to reconstruct the earliest Germanic meaning of nes- as “zum Leben zurückkehren (return to life).” [1] The correlation between this and what has been found in early Greek leads to the conclusion that the meaning “return to life” goes back to Indo-European times.
The discussion to follow will deal first with the forms from nes- in Gothic, secondly with those in West Germanic, and lastly with those in Nordic. In each case the original meaning “zum Leben zurückkehren” can still be seen, or at least inferred. In each case there is also a certain amount of evidence for an original meaning “return to light,” although the nature and strength of this evidence vary from dialect to dialect.
In dealing with Gothic one must take account of the problem that our evidence is contained in a translation of the Greek Bible. The context in which words appear is therefore predetermined. The element of choice, however, enters into the translator’s use of a particular word in a given context. In the case of the root nes-, this point has a kind of general relevance; for the Gothic forms from this root are used to translate such Greek words as sṓzein and sōtēría, which in the original text have to do with the specifically Christian notion of “salvation.” This notion, furthermore, implies a “return from death” to life everlasting. While it does not follow automatically from this that the Gothic forms from nes- imply a “return from death,” this possibility is at least distinctly present.
The same argument holds for the meaning “return to light.” At the center of Christianity there is a sōtḗr, “savior,” who is {126|127} called not only “the life,” but also “the light.” [2] The Gothic translation of sōtḗr is nasjands, and as an example of a significant context in which the word appears, I quote the Greek original of 2 Tim. 1:10, which speaks of the “grace” conferred by Christ:
(χάριν) φανερωθεῖσαν δὲ νῦν διὰ τῆς ἐπιφανείας τοῦ σωρῆρος ἡμῶν (nasjandis unsaris) Ἰησοῦ Χριστοῦ, καταργήσαντος μὲν τὸν θάνατον, φωτίσαντος δὲ ζωὴν καὶ ἀφθαρσίαν διὰ τοῦ εὐαγγελίου
But it [grace] is now made manifest by the illumination of our savior Jesus Christ, who has nullified death, and has brought to light life and incorruption through the gospel. [3]
In the following I give the various Gothic forms and a sample context in which each appears. Most of the total occurrences of each form refer to Christian redemption, for which much is implied, but I will not limit myself to these, for other contexts as well can be revealing. [4]
The verb ganisan, which translates Greek sṓzesthai, “be saved,” is intransitive and is composed of the root nes- in its Gothic form together with the perfective element ga-. [5] The verb demonstrates that the meaning “get well” of its modern German equivalent, genesen, is old; cf. Mark 5:23: ἵνα σωθῇ καὶ ζήσῃ, ei ganisai jah libai, “that she may get well and live.” {127|128} For the context of Christian salvation, cf. 1 Cor. 5:5:
παραδοῦναι τὸν τοιοῦτον τῷ Σατανᾷ εἰς ὄλεθρον τῆς σαρκός, ἵνα τὸ πνεῦμα σωθῇ (ei ahma ganisai) ἐν τῇ ἡμέρᾳ τοῦ κυρίου Ἰησοῦ
To deliver such a one to Satan for the destruction of the flesh, that the spirit may be saved in the day of our lord Jesus.
Based on ganisan is the feminine noun ganists, which is formed with the abstract suffix -ti-, and which translates sōtēría. [6] Christian “salvation” is opposed to death in the following context, 2 Cor. 7:10:
ἡ γὰρ κατὰ θεὸν λύπη μετάνοιαν εἰς σωτηρίαν (du ganistai) ἀμεταμέλητον κατεργάζεται, ἡ δὲ τοῦ κόσμου λύπη θάνατον κατεργάζεται.
For sorrow which is according to God brings about repentance, steadfast unto salvation; but the sorrow of the world brings about death.
The verb nasjan is a causative formation and therefore has a transitive sense, translating sṓzein, “to save.” Christian salvation is again opposed to death in the following phrase, Mark 3:4: ψυχὴν σῶσαι (saiwala nasjan) ἢ ἀποκτεῖναι, “to save a soul or kill it.” The participle from nasjan is nasjands, “sōtḗr,” which was discussed above.
The verb nasjan also appears as ganasjan, with the perfective element ga-. This verb translates sṓzesthai, “be saved,” in the following interesting example; in Luke 8:50 the context involves a young girl presumed to be dead, concerning whom Christ says to her father: μόνον πίστευε, καὶ σωθήσεται (jah ganasjada), “only believe and she will be saved.” The verb also translates iãsthai, “to heal,” and thus acts as a causative to ganisan in the sense “get well” of the latter; cf. Luke 6:19: καὶ ἰᾶτο πάντας (jah ganasida allans), “and he healed all.” {128|129}
Based on the verb nasjan is the feminine noun naseins, formed with the suffix –eins from original *-ī-ni-, and translating sōtēría and sōtḗrion; the latter stands in the original of Luke 2:30 ff.:
ὅτι εἶδον οἱ ὀφθαλμοί μου τὸ σωτήριόν σου (nasein þeina) ὃ ἡτοίμασας κατὰ πρόσωπον πάντων τῶν λαῶν, φῶς εἰς ἀποκάλυψιν ἐθνῶν
Because my eyes have seen your salvation, which you have prepared before the face of all people, a light for the revelation of the Gentiles.
This completes the list of Gothic forms, the contexts of which must speak for themselves, and their cognates in West Germanic may now be considered. Here the meanings of Gothic ganisan, both “be saved” and “get well,” are closely reflected in Old English genesan (also nesan) and Old Saxon ginesan, “be saved,” and Old High German genesan, “be saved, get well”; similarly, Old Saxon and Old High German ginist means “salvation” and “recovery.”
The cognates of Gothic nasjan require more comment, for here there was, in part, a semantic development, reflected in the modern German cognate nähren, “nourish.” Nasjan itself simply means “save,” and this meaning was preserved in Old English and Old Saxon nerian; similarly, the participles neriend and neriand mean “savior” in the two respective dialects.
Before proceeding to Old High German, it is worth giving one example of Old English nerian, “save,” in which the sense “bring back to light” is at least implied by the context. The example occurs in Beowulf 569 ff., a passage in which the hero tells of the aftermath to his victory over the “nickers nine,” the sea-monsters he overcame one night in the ocean depths:
Leoht eástan com,       Light came from the east,
beorht beácen Godes,       God’s bright beacon,
brimu sweþrodon,       the seas grew calm,
þæt ic sǽ-næssas       so that the sea-nesses I
geseón mihte,       might see,
windige weallas.       windy walls.{129|130}
Wyrd oft neređ       Fate often saves
unfǽgne eorl,       an undoomed man,
þonne his ellen deáh.       when his valor avails.
“Salvation” in these lines, with their emphasis on the sun and the hero’s ability to see about him, seems intimately related to a “return to light.” [7]
In the Old High German forms nerian and nerren, the meaning “nourish” developed, as also in Old Frisian nera; alongside this new meaning, the old meaning “save” persisted even into modern High German, but today only the developed meaning “nourish” is current. [8] For the development of this new meaning, one might compare the Latin idiom corpus curare cibo.
The semantic development in Old High German and Old Frisian is of importance for the Nordic evidence, to which we now come. For it seems likely that the Nordic languages borrowed the German word in its developed meaning, and that this explains the origin of the following forms, all meaning “nourish”: Old and Modern Icelandic næra, Norwegian næra, modern Swedish nära, and Danish nære. [9]
From these forms, which are found in both East and West Norse, should be distinguished another set of forms, which are limited to West Norse. The latter forms appear to be inherited, and it is the meaning “kindle” of two of them in which we are interested. This is the meaning of Norwegian and Swedish-dialect nöra. [10] Beside the meaning “kindle” in Norwegian and Swedish- {130|131} dialect, the meanings “stärken, erfrischen, ernähren (invigorate, refresh, nourish)” are attested in Old Icelandic nœra, modern Icelandic nœra, and Norwegian dialect nöra. [11]
Thus, West Norse appears to have had a verb with two separate meanings, “revive” and “kindle.” The formal evidence suggests by its distribution that the meaning “kindle” developed from a more basic meaning of the verb. But then the problem is to determine precisely what this basic meaning was. I do not agree with the suggestion [12] that the meaning “kindle” developed from the notion of “feeding” or “nourishing” a fire. There are two arguments against this. The first is simply semantic; it is difficult to see how the expression “feed a fire” could come to mean “light a fire.” But the second argument is more basic; it is not the set of forms represented by nöra, “kindle,” that has “nourish” as its primary meaning. This is, rather, the set of forms apparently borrowed from West Germanic. [13]
In order to determine the oldest meaning of the first set, which is native to West Norse, one must consider the evidence of Old Icelandic. It emerges from this that Old Icelandic nœra did not mean “nourish” at all, but still had the older sense of “revive, bring back to life.”
For the sake of comparison, let us first cite an example of Old Icelandic næra, “nourish.” These occurrences are few and late—thus the phrase næra sín hibýli, “to feed his family,” in the Konungs Skugg-sja (326B), a late translation of the Speculum Regale.
The meaning of nœra is quite distinct from this, as the following example will show. (In these examples I shall quote the spelling of the texts, which frequently have næra for nœra; {131|132} the reason for this is that -æ- and -œ- have fallen together in modern Icelandic, thus obscuring the distinction in which we are interested; the meanings, however, will serve to identify the forms as coming from nœra.) An example that might at first seem to have to do with “feeding” but does not, is found in Fornaldar Sögur 3.571: hón dreypir víni á varrir þeim ok nærdusk þeir skjótt, “she dripped wine on their lips, and they brought themselves back to life again.” The meaning naturally cannot be “and they nourished themselves again.”
For nœra as “bring back to life,” one might also consider the following striking example, Fornmanna Sögur 6.353: nærdisk hón svá sem frá leid, “she brought herself back to life who had perished (i.e., swooned)”, where the context has nothing at all to do with “nourishing.”
These examples serve to show that the meaning “kindle” of Norwegian and Swedish dialect nöra has to do, not with “feeding,” but with “bringing back to life.” The association with fire is important, because it suggests that West Norse could have inherited the meaning “return to light” as well as “return to life” from the root nes-. The association with fire, furthermore, is old, to judge by the following Old Icelandic passage, Fornmanna Sögur 10.368; this passage concerns a queen who says that she has been so wasted by grief for her husband, at engi gneisti lífsins má nœra mik edr lífga, “‘that no spark (geneisti) of life may “kindle” me or bring me back to life.’” [14]

3. Albanian

One of the forms commonly cited as a derivative from Indo-European nes- is the Albanian verb kneƚem, “erhole mich, werde {132|133} wieder lebendig (I recover, come back to life).” [15] The derivation was first proposed by N. Jokl, [16] who saw in the meaning of this verb a clear correspondence to the Germanic forms discussed above. The verbal base of kneƚem is to be segmented k-nel-, with k- a perfectivizing element like Germanic ga-. The element -nel- would then go back to *nes-l-, and the verb itself would derive from an adjective built on a formant -lo-.
If this derivation is correct, as the meaning of kneƚem would argue, then it provides a useful piece of evidence. For “werde wieder lebendig” is only one of the meanings of the verb. For the total range of the verb, I quote the complete entry contained in a recent Albanian-English historical dictionary. [17] Used intransitively, the verb means: “recover, get well; thrive; (of light) brighten.” Used transitively, it means: “refresh, revive; make red-hot.”
Jokl was aware of the meaning “brighten” from an Italian dictionary of northern Albanian, which glossed kneƚem with “divenire vivo e splendente, rischiararsi,” and he drew attention to Scandinavian nöra, “kindle.” He assumed that the similar meanings in the two cases resulted from an independent “pregnant” usage. [18] Based on the evidence of Greek, however, it seems more than likely that the Albanian “prägnante Verwendung für das Feuer oder die Sonne (pregnant usage for fire or the sun)” has been inherited from the same Indo-European source as the meaning “kindle” of Scandinavian nöra. {133|134}

4. Indic

Sanskrit násate is an exact formal cognate to Greek néomai. But the meanings of the Sanskrit verb, “approach, resort to, join,” have developed too far to shed light on the meanings of the Indo-European verb. [19]
Both Sanskrit and Avestan have a neuter noun astam, “home,” which has been reconstructed as *n̥s-to-, with zero-grade of the root nes-. [20] This reconstruction is probable, and it indicates that the secular meaning “to return home” of Greek néomai goes back to Indo-European (see above chap. 2, text preceding n. 24).
For the sacred meaning of IE nes-, the crucial Sanskrit form is the proper name Nā́satyā. This name, which is in the dual, belongs to a pair of twin gods in the Vedic pantheon. The etymology of Nā́satyā has yet to be established with certainty, but the prevailing modern view is to derive the form from the root nes-. [21] The Vedic twins have the characteristic functions of “saving” and “healing” distressed mortals, and this has suggested a connection with the Gothic verb nasjan, “to save” and “to heal.” In the rest of n̥form contains precious comparative evidence for the sacred meaning of the Indo-European root.
Sanskrit Nā́satyā has equivalents in Avestan and Mitanni, and is thus known to go back to Common Indo-Iranian. Avestan Nā̊ŋhaiθya, a singular form, designates a demon in the Zoroastrian system, and Mitanni Nasattii̪a(nna), a plural form, apparently designates the same twin gods as Vedic Nā́satyā. [22] {134|135}
Since Sanskrit Nā́satyā originated at least as early as Common Indo-Iranian, it is probable that even the Vedic poets no longer knew what it meant. Later Indian tradition provides three distinct interpretations of the name, but all are based on folk-etymologies. One of these ( = nar + satya, “true warriors”) is clearly impossible; [23] a second, found in Pāṇini (na-asatya, “not untrue”) has rightly been rejected on both formal and semantic grounds; [24] the third (nāsā-tya, “the nose-born”), although it has been defended in modern times by Herman Lommel, is equally to be rejected. [25]
The first to suggest a derivation from the root nes- was Brunnhofer. [26] The difficulty, however, has been to explain the morphology of the form, and for this Hermann Güntert is {135|136} credited with the best solution. [27] Güntert began with the Indo-Iranian ancestor of Sanskrit násate, to which he ascribed the meaning “hasten to safety” on the basis of the Germanic evidence. He then posited a nominal derivative *nasati-, “salvation through hastening to the rescue,” on the parallel of vasatí-, “dwelling,” derived from vásati, “to dwell.” The name Nā́satyā would then be based on *nasati- just as the name Ādityá- is based on áditi-.
Güntert’s solution is possible, but the hypothetical abstact noun *nasati- is weak. One would have expected *nasti- in view of Gothic ga-nists, or even *asti- on the basis of Indo-Iranian astam. Given this weakness, I would propose a different solution.
Gregory Nagy, in a discussion of Old Persian xšāyaθiya-, “king,” has argued that the underlying form *kšāyatya- is perhaps a “nominalization (with vṛddhi) of an archaic syntagma *kšayati-ya, ‘he who has power.’” [28] Such a combination of third-person verb plus relative pronoun has parallels in Celtic and perhaps Italic, and the syntactic order is well attested in Indo-Iranian. [29] If Nagy is correct about the origins of *kšāyatya-, I propose that Indo-Iranian Nā́satya- is likewise a nominalization (with vṛddhi) of an archaic syntagma *nasati-ya, “he who brings back to life and light.” The verb *nasati would be an activized {136|137} form, otherwise unattested, of násate, and as such would be parallel to the Homeric verb *nései that we have reconstructed in xviii 265. [30]
A singular verb *nasati would imply that Nā́satyā was also originally singular. Avestan Nā̊ŋhaiθya provides comparative evidence for such a singular in Common Indo-Iranian, and even the Rig-Veda contains one instance of a singular “traversing Nā́satya-” in 4.3.6. [31] For the most part, however, the Rig-Veda does not make distinctions between the two twins but treats them as an identical pair. Hence we must begin with the notion of two Nā́satyā, who both “bring back to life and light.”
The two twins, as noted above, “save” and “heal” distressed mortals. They perform these functions in a series of archaic myths which the Rig-Veda refers to frequently. Three such myths, furthermore, provide explicit evidence that the underlying function of the two “saviors” and “healers” was in fact “to bring back to life.” The myths concern the mortals Rebha, Bhujyu, and Śyāva. {137|138}
Rebha was bound, stabbed, and cast into the waters for nine days and ten nights before being saved by the twins. RV 10.39.9 says that he was “dead” (mamṛvā́ṃsam) when the twins “raised him up” (úd airayatam). [32]
Bhujyu was saved after his father or evil companions abandoned him at sea. RV 1.119.4 refers to the twins as “bringing (Bhujyu) home from the dead ancestors” (niváhantā pitṛ́bhya ā́). [33]
Little is known about Śyāva, to whom there are only two references in the Rig-Veda. [34] But one of these, RV 1.117.24, says that he was “split in three” (trídhā … víkastam) when the twins “raised him up to live” (új jīvása airayatam).
Not only do the twins “bring back to life,” but they also “bring back to the light.” To Bhujyu, who was “cast forth on the unsupporting darkness” (anārambhaṇé támasi práviddham, 1.182.6), the twins gave “light-bringing help” (svàrvatīr … ūtī́r, 1.119.8). Rebha they “raised up to see the sun” (úd … aírayataṃ svàr dṛśé, 1.112.5), and likewise Vandana (ibid.). Vandana, who was buried in a pit when he disintegrated with old age, is also the object of a pair of similes which strikingly connect the ideas of “returning from death” and “returning from darkness.” RV 1.117.5 says that the twins restored Vandana “like one who had fallen asleep in the womb of the death goddess” [35] (suṣupvā́ṃsaṃ {138|139} ná nírṛter upásthe), and “like the sun dwelling in darkness” (sū́ryaṃ ná … támasi kṣiyántam).
The twins, who were invoked at dawn, have a close connection with Uṣas, the dawn goddess. She is bidden to awaken them (8.9.17), they follow her in their chariot (8.5.2, etc.), they hitch their steeds when she is born (10.39.12), and their chariot is once said to arrive before her (1.34.10). [36] As their hour of invocation indicates, the twins have to do with a “return from darkness” in cult as well as myth. In three hymns the poet marks the time of invocation with the phrase “we have reached the other shore of this darkness” (átāriṣma támasas pārám asyá). [37] In another hymn the poet prays to the twins for “refreshment” (íṣam) “which will deliver us across the darkness” (yā́ naḥ pī́parad … támas tiráḥ, 1.46.6). The twins are called “darkness slayers” (tamohánā, 3.39.3), and their horses and chariot are described as “uncovering the covered darkness” (aporṇuvántas táma ā́ párīvṛtam, 4.45.2). One hymn invokes the twins as “you who have made light for mankind” (yā́v … jyótir jánāya cakráthuḥ, 1.92.17).
The Vedic twins, then, clearly have the function of “bringing back to life and light” in a context of solar mythology. Their connection with solar mythology, moreover, is known to go back to Indo-European. The twins are closely associated with a female figure named Sūryā́, a feminized form of Sū́rya, the “Sun.” The twins are called Sūryā́’s husbands and wooers, and she is frequently said to mount their chariot at dawn. [38] Sūryā́ is also called duhitā́ sū́ryasya, “the daughter of the Sun,” [39] and this name corresponds exactly to Lithuanian Saulės dukterys and {139|140} closely to Latvian Saules meita. These two “daughters of the Sun” are associated with the Baltic counterparts to the Vedic twins, the Lithuanian Dievo sunėliai and the Latvian Dieva dēli. In Vedic the “daughter of the Sun” is the common wife of the two twins, while in Baltic she is their sister. In Indo-European she was probably both wife and sister simultaneously. [40]
Now that I have made a case for the two Nā́satyā as “they who bring back to life and light,” I return to the question of an originally singular form. I shall begin by considering comparative evidence having to do with the Greek Dioscuri. Like the Vedic twins, the Dioscuri are “saviors” and “healers” of distressed mortals, and their cult title and literary epithet, sōtē̂res, “saviors,” has been compared with the Vedic name Nā́satyā. [41] Unlike the Vedic twins, however, the Dioscuri have preserved their internal oppositions, and from this perspective, only one of the twins is a “savior.” Since Polydeuces is immortal and Castor mortal, Polydeuces actually brings Castor “back to life.”
The crucial moment for the Dioscuri is their battle with Idas and Lynceus, when Castor is mortally wounded and Polydeuces gives up half of his immortality in order to save him. Our best source for the episode is Pindar’s Nemean 10. [42] In his introduction to the narrative, Pindar says that Polydeuces, given the choice of becoming entirely a god, instead “chose the life of Castor who had perished in battle” (l. 59):
εἵλετ’ αἰῶνα φθιμένου Πολυδεύκης Κάστορος ἐν πολέμῳ. {140|141}
In the narrative itself, Pindar says that Polydeuces “opened the eye, then the voice” of his fallen brother (l. 90):
ἀνὰ δ’ ἔλυσεν μὲν ὀφθαλμόν, ἔπειτα δὲ φωνὰν χαλκομίτρα Κάστορος.
In Pindar’s account, the immortal Polydeuces clearly brings his mortal brother “back to life.” [43]
We cannot know for sure whether the immortal/mortal opposition which determines the actions of the Greek twins also characterized the Vedic twins, but this is a likely supposition. The Greek twins are so opposed because they have different fathers, {141|142} the immortal Zeus and the mortal Tyndareus. [44] The Vedic twins also have different fathers, and one of these is Dyáus, the exact cognate of the Greek Zeús. This information comes from RV 1.181.4, the only Vedic text which explicitly distinguishes one twin from the other. The same text calls the other father Súmakha, “Good Warrior,” a figure who is unknown otherwise but who was almost certainly mortal like the Greek Tyndareus. The dual paternity of the Greek and the Vedic twins apparently goes back to the same Indo-European souce, where one of the fathers was the immortal “Sky God” *Dyēus, and the other father, in all likelihood, a mere mortal. [45] If Sumakha was indeed mortal, then his son must also have been mortal, and it is hard to imagine how such a “mortal” twin could continue to exist unless his immortal brother brought him “back to life.”
As stated earlier, the Vedic twins are closely connected with sunrise in the Rig-Veda. But in addition to their well-attested morning ritual, the twins also had an evening ritual. The two times, evening and morning, are clearly paired and contrasted in three hymns of the Rig-Veda: doṣā́ … uṣási in 8.22.14, doṣā́m uṣā́saḥ in 10.39.1, and doṣā́ vástor in 10.40.4. [46] The evening/morning opposition which characterizes the twins’ ritual must also have characterized the twins themselves. A passage quoted by Yāska (Nirukta 12.2) bears this conclusion out by stating that “one (of the twins) is called the son of night, the other the son of {142|143} dawn.” [47] Now if only one twin was properly connected with sunrise, then only one twin can properly have had the function of “bringing back to life and light.” It is this twin who is to be connected with the name Nā́satya-.
To continue this investigation I now shift my attention from pre-Vedic to post-Vedic tradition. The Swedish scholar Stig Wikander has brilliantly shown that Sanskrit epic indirectly preserves old oppositions between the twins which the Rig-Veda ignores. [48] Two of the heroes of the Mahābhārata, the twins Nakula and Sahadeva, are depicted as sons of the divine twins, and, as Wikander has shown, are characterized in terms of their fathers’ archaic attributes. [49] But while Nakula and Sahadeva often act in common like their Vedic fathers, they are not an identical pair: rather, they are opposed to each other as “warrior horseman” to “intelligent cattleman.” These oppositions can be shown to be old, and Sahadeva’s attributes—both “cattle” and “intelligence”—are of great importance to the name Nā́satya-. {143|144}
Wikander began his study of the epic twins with a survey of their characteristic epithets. What distinguishes Sahadeva absolutely from his brother is his characterization as “intelligent”: [50] he is called “wise” (vidvān, 17.2.54), “intelligent,” (prājña, 17.2.56), “endowed with understanding” (buddhimān, 14.72.2103), “learned” (paṇḍita, 2.63. 2155), “clever” (matimān, 3.269.15710), “acute” (nipuṇa, 5.49.1838), and “clairvoyant” (cakṣuṣin, 6.75.3282).
Nakula, on the other hand, is “warlike” and possessed of a warrior’s “beauty.” [51] He is called “skillful in all forms of war” (sarvayuddhaviśārada, 7.165.7364), “good in war” (kuśalaṃ yuddhe, 7.98.3976), and, most characteristically, “fighting in a wondrous manner” (citrayodhin). [52] He is also called “beautiful” (darśanīya, 3.27.1020, 4.3.61, 5.49.1996), “the most beautiful in the world” (darśanīyatamo loke, 2.78.2625), and “the most beautiful of heroes” (darśanīyatamo nṛṇām, 2.75.2555).
Wikander discussed two scenes in the Mahābhārata which underscore the opposition between Nakula’s “warrior beauty” and Sahadeva’s “intelligence.” [53] In the great dice game, Yudhiṣṭhira, the twins’ eldest brother, characterizes both of them {144|145} when he puts them up as stakes: Nakuka he calls “dark, young, with eyes of flame, the shoulders of a lion, and huge arms” (2.63.2152), while Sahedeva “teaches justice and has acquired in this world a reputation for being learned” (paṇḍita, 2.63.2155). Similarly, near the end of the poem, when the twins die, Yudhiṣṭhira interprets their deaths as the result of peculiar moral flaws: [54] Nakula’s flaw was to think “there is no one equal to me in beauty” (rūpeṇa matsamo nāsti kaścid iti, 17.2.62); Sahadeva, on the other hand, “always thought that no one was as intelligent as himself” (ātmanaḥ sadṛśaṃ prājñaṃ naiṣo ’manyata kañcana, 17.2.56). [55]
The second great opposition between the epic twins contrasts “horses” with “cattle.” Wikander discovered this opposition in the fourth book of the Mahābhārata, where the twins and their three older brothers all assume different disguises in order to spend their last year of exile at the court of king Virāṭa. The twins both disguise themselves as vaiśya-s—members of the third caste, having to do with agriculture and production. This disguise {145|146} is a reflection of their fathers’ archaic nature, for the divine twins had once been the vaiśya gods. [56] The disguise as vaiśya-s, however, also reveals an opposition between the twins that would not have been suspected otherwise: while Nakula disguises himself as a groom and takes charge of Virāṭa’s “horses,” Sahadeva speaks warmly of his preference for “cattle” and becomes Virāṭa’s cowherd. [57] This opposition between horses and cattle has been shown by Dumézil to be as old as Indo-European, [58] and it will prove crucial in my further analysis of the Vedic twins.
To correlate the oppositions between Nakula and Sahadeva with the Vedic twins, we begin with RV 1.181.4, which calls one twin the son of Dyaus and the other the son of Sumakha. The son of Súmakha, “Good Warrior,” is also called “conquering” (jiṣṇú) and “lordly” (sūrí), and, as Wikander has argued, [59] he plainly corresponds to the “warrior” Nakula. Sahadeva, then, corresponds to the son of Dyaus.
The crucial point, however, is to connect Sahadeva, the “intelligent cattleman,” with the name Nā́satyā. To do this we now take into account the fact that the Vedic twins actually have two names, both in the dual: they are not only the Nā́satyā, but also the Aśvínā, the “Horse-Possessors.” As I will argue below, the Vedic names Nā́satyā and Aśvínā function like the Greek names Dióskouroi, “Sons of Zeus,” and Tundarídai, “Sons of Tyndareus”: [60] each name refers properly to a different twin. {146|147} Further, the name Aśvinā, “Horse-Possessors,” corresponds to the “horseman” Nakula, and the name Nā́satyā corresponds to Sahadeva, the “intelligent cattleman.”
Wikander has also argued that the two dual names were originally singular, but he thought that both names belonged to the same twin, the father of Nakula. [61] Wikander’s own methodology, however, supports a different conclusion.
Following a suggestion of Geldner’s, Wikander noticed that certain of the twins’ dual epithets have a statistical tendency to occur in the same strophe with the dual name Aśvínā. [62] Such co-occurrences, he argued, reveal old oppositions between the epithets in question and the name. [63] One of the epithets is divó nápātā, “sons of Dyaus,” which, like Greek Dióskouroi, represents the extension of one twin’s title to both twins. In three of the five occurrences of divó nápātā in the Rig-Veda, the name Aśvína occurs in the same strophe. The ratio 3:5 bears out an opposition between the “son of Dyaus” and the father of the “horseman” Nakula which we have already seen indicated.
Wikander detected two further oppositions by this method. An opposition between Aśvínā and vṛ́ṣaṇā, “bulls,” is indicated by a ratio of 13:28, and an opposition between Aśvínā and dasrā́, “the miracle-workers,” in indicated by a ratio of 24:44. [64] Wikander failed to observe, however, that a comparable ratio of 32:100 indicates an opposition between Aśvínā and Nā́satyā.
More important than the statistical count, however, is the evidence of one strophe which contains the Nā́satyā/Aśvínā oppositions: in RV 2.41.7 this opposition is closely correlated with a further opposition between “cattle” and “horses.” The {147|148} strophe is divided into three segments, the first two of which articulate the two oppositions in question. The first segment begins with the adjective gómad, “rich in cattle,” and ends with the vocative Nāsatyā, which the second segment begins with the adjective áśvāvad, “rich in horses,” and ends with the vocative Aśvinā. The parallelism between these two segments makes it clear beyond doubt that “cattle” are associated with the name Nā́satyā and “horses” with the name Aśvínā:
gómad ū ṣú nāsatyā/ áśvāvad yātam aśvinā/ vartī́ rudrā nṛpā́yyam
Come quickly (ū ṣú … yātam) along the path (vartī́) rich in cattle, you Nāsatyā; come quickly along the path rich in horses, you Aśvinā; come quickly, you Rudrā, along the path which protects men (nṛpā́yyam).
The evidence of RV 2.41.7, decisive in itself, [65] is further confirmed by RV 7.72.1. This strophe has two segments, and the adjectives “rich in cattle” (gómatā) and “rich in horses” (áśvāvatā) are again set in opposition to one another in different segments. Although the name Aśvínā is omitted in this strophe, the name Nā́satyā is again plainly associated with “cattle” in segment 1:
ā́ gómatā nāsatyā ráthena / áśvāvatā puruścandréṇa yātam
Come here (ā́ … yātam) on your chariot (ráthena) rich in cattle, you Nāsatyā; come here on your chariot rich in horses and abundant with gold (puruścandréṇa).
The evidence of RV 2.41.7, supported by that of 7.72.1, establishes conclusively that the names Nā́satyā and Aśvínā {148|149} originally designated different twins, and that the name Nā́satyā is to be correlated with the “cattleman” Sahadeva, while the name Aśvínā is to be correlated with the “horseman” Nakula. It is also now certain that the son of Dyaus, the prototype of Sahadeva, had the name Nā́satya- in opposition to his brother. There is thus every reason to believe that this Nā́satya-, like the Greek son of Zeus, originally brought his brother “back to life and light.” The “cattle” and “intelligence” of Sahadeva establish beyond reasonable doubt that the name Nā́satya- derives from the root nes-, and that it originally signified “he who brings back to life and light.” [66]
We now return to take account of Wikander’s argument that a singular Nā́satya- originally designated the father of Nakula. Wikander based this argument on the dual compound name Nāsatyadasrau, which is applied to the divine twins in Sanskrit epic, and which names one of the twins Nā́satya- and the other Dasra-. [67] The name Dasra- comes from the twins’ dual epithet dasrā́, “miracle working,” which Wikander’s statistical argument showed to be opposed to the name Aśvínā (see earlier in this section). {149|150} Since the epic compound opposes dasrā́ to the name Nā́satyā, Wikander concluded that the names Aśvínā and Nā́satyā originally designated the same twin.
The problem with Wikander’s argument is his reliance on the epic compound, which cannot be old. Wikander himself admitted that statistics do not bear out an opposition between the duals Nā́satyā and dasrā́ in the Rig-Veda. [68] The two terms, in fact, have a tendency not to occur in the same verse. [69] Once again, however, specific texts are more conclusive than statistics. There are three more strophes in the Rig-Veda that contain an opposition between “cattle” and “horses,” and in these strophes dasrā́ actually takes the place of Nā́satyā on the “cattle” side of the opposition. This shows as clearly as possible that the terms {150|151} Nā́satyā and dasrā́, far from being opposed, are in fact isofunctional, and that the compound Nāsatyadasrau is secondary.
All three strophes contain the phrase gómad dasrā híraṇyavat, consisting of the vocative dasrā and the neuter adjectives gómad, “rich in cattle,” and híraṇyavat, “rich in gold.” In RV 8.22.17, the collocation gómad dasrā in the third segment of the strophe is clearly opposed to the collocation áśvāvad aśvinā in the first segment:
ā́ no áśvāvad aśvinā / vartír yāsiṣṭam madhupātamā narā / gómad dasrā híraṇyavat
Come to us (ā́ no … yāsiṣṭam) along the path (vartír) rich in horses, you Aśvinā; come, you who most enjoy the sacrificial drink, you heroes (madhupātamā narā); come along the path rich in cattle, you dasrā, rich in gold.
RV 1.30.17 contains the same opposition, but the syntax is varied on the “horse” side of this opposition, and the phrase gómad dasrā híraṇyavat is appended loosely to express the “cattle” side:
ā́śvināv áśvāvatyā / iṣā́ yātaṃ śávīrayā / gómad dasrā híraṇyavat
Come here (ā́ … yātam) with a surpassing wealth (iṣā́ … śávīrayā) rich in horses, you Aśvinā; rich in cattle, you dasrā, rich in gold.
RV 1.92.16 follows the pattern of the two previous strophes but omits the adjective áśvāvad altogether:
áśvinā vartír asmád ā́ / gómad dasrā híraṇyavat / arvā́g ráthaṃ sámanasā ní yachatam
Being of one mind rein in your chariot in this direction (segment 3), along the path toward us, you Aśvinā (segment 1), which is rich in cattle, you dasrā, and rich in gold (segment 2). {151|152}
By comparing the three texts just examined with the two that were examined earlier (RV 2.41.7 and 7.72.1), we may conlude that the terms Nā́satyā and dasrā́ are indeed isofunctional. This conclusion, furthermore, provides a final, important piece of evidence for the etymology of Nā́satyā. The epithet dasrā́, as stated, means “miracle-working,” and it relates to the twins’ function as “magic healers.” [70] Since the epithet is isofunctional with the name Nā́satyā, there is now every reason to connect this name with the root nes-, and with the “miraculous” function of “bringing back to life and light.”
The epithet dasrā́ also confirms that the name Nā́satyā has to do with “intelligence.” The epithet has an exact cognate in Avestan daŋrō, which means “clever, skillful,” and the Indo-Iranian forms are in turn related to the family of Greek daē̂nai, “to learn.” Sanskrit dasrá must also have implied “intelligence,” [71] and this quality, which is so essential to the epic Sahadeva, is thus confirmed for his Vedic prototype as well.
It is at least highly probable that the name Nā́satyā, as derived from the root nes-, implied “intelligence” directly. The name is thus a precious comparison for Greek nóos, indicating that the root nes- had to do with “intelligence” already in Indo-European. [72] {152|153}


[ back ] 1. S. Feist, Vergleichendes Wörterbuch der gotischen Sprache 3 (Leiden, 1939), s.v. ganisan.
[ back ] 2. Cf., e.g., John 8:12: “I am the light of the world: he that followeth me shall not walk in darkness, but shall have the light of life.”
[ back ] 3. It may be noted that the image of Christ as tò phō̂s, “the light,” completely dominates the fourth Gospel, the one which betrays the greatest Greek influence; cf. A Dictionary of Christ and the Gospels, vol. 2, ed. J. Hastings (Edinburgh, 1924), pp. 34–35. A point unrelated to Gothic, but in itself well worth noting, is that in Greek Hermetic writers there are about ten references to a deity called Noũs and defined as phō̂s, “light,” and zōḗ, “life”; for these references, see W. Arndt and F. W. Gingrich, A Greek-English Lexicon of the New Testament and Other Early Christian Literature (Chicago, 1957), p. 341.
[ back ] 4. For a list of attestations of the various Gothic forms, see H. C. von der Gabelentz, Ulfilas, vol. 2 (Leipzig, 1843), s.vv.
[ back ] 5. For the derivations given here, see Feist (n. 1), s.vv.
[ back ] 6. Cf. chap. 4, n. 60 for the possibility of a similar Greek formation *néstis.
[ back ] 7. Rhys Carpenter, Folktale, Fiction, and Saga in the Homeric Epics (Berkeley, Calif., 1956), pp. 138–139, compares Beowulf’s descent to the ocean depths to Odysseus’s descent to the underworld, arguing that both derive from the same ultimate source (the “Bearson” legend); the argument is at least interesting in light of the posited etymological meaning of neređ as “brings back from death.”
[ back ] 8. Cf. J. and W. Grimm, Deutsches Wörterbuch IV.1.2.7 (Leipzig, 1878), s.v. genesen.
[ back ] 9. See J. De Vries, Altnordisches etymologisches Wörterbuch (Leiden, 1962), s.v. næra, who further argues that the borrowings were not from Old High German, but rather from a Low German form neren (which one must assume had the same meaning).
[ back ] 10. The Swedish dialects in question are those of Bohuslän and Dalsland, both of which are West Norse in origin.
[ back ] 11. For Icelandic, cf. De Vries (n. 9), s.v., and A. Noreen, Altisländisches Grammatik (Halle, 1890), p. 90; for Norwegian dialect, cf. I. Aasen, Norsk Ordbog (Christiania, 1873), s.v.
[ back ] 12. Made by R. Meringer, Wörter und Sachen 1 (1909): 168 ff.
[ back ] 13. It should be noted that Faroese nøra, “provide with food; breed or raise (livestock)” does not contrast with another form *næra; has the falling together of two forms thus given nøra the meaning “nourish”? Such a coincidence of forms has taken place in the case of Icelandic (see below in text).
[ back ] 14. It should also be noted that corresponding to the Old English compound ealdorneru, “life’s salvation, refuge,” is the Old Icelandic compound aldrnari, which, however, is a designation for “fire” in the context of the Ragnarök—the destruction and recreation of the world (Völuspa 57.3).
[ back ] 15. Cf. J. Pokorny, Indogermanisches etymologisches Wörterbuch (Bern, 1954), 2:766; M. Mayrhofer, Kurzgefasstes etymologisches Wörterbuch des Altindischen (Heidelberg, 1956–1974), s.v. nasate; H. Frisk (chap. 1, n. 4), s.v. néomai. (Frisk, however, adds a caution that the form is ambiguous.)
[ back ] 16. N. Jokl, Sitzungsberichte der Akademie der Wissenschaften zu Wien 168: 1, 40.
[ back ] 17. Stuart E. Mann, An Historical Albanian-English Dictionary (Cambridge, 1957).
[ back ] 18. Jokl repeats the notion that the basic meaning of Old Icelandic nœra is “ernähren (nourish)”; but one should note that at least Albanian kneƚem has nothing to do with “feeding.”
[ back ] 19. Sanskrit also has the reduplicated form niṃsate, “kiss,” to which Greek nísomai is thought to correspond; the difficulty with the Greek form is that original -ns- should have produced -n-, not -s-. For a bibliography of proposed solutions, see Frisk (chap. 1, n. 4), s.v. néomai.
[ back ] 20. For the derivation, see M. Mayrhofer (n. 15), s.v.
[ back ] 21. See n. 27 below.
[ back ] 22. The Mitanni name occurs on a treaty from the fourteenth century B.C. between the Mitanni king Matiwāza and the Hittite king Suppiluliuma. Mitanni equivalents of the Vedic Mitra, Varuṇa, and Indra occur on the same treaty and thus establish the identity of the Nasattii̪a(nna). For the Mitanni evidence, see M. Mayrhofer, Die Indo-Arier im alten Vorderasien (Wiesbaden, 1966), pp. 14–15, with further bibliography; there is a convenient summary of the evidence in T. Burrow, The Sanskrit Language 2 (London, 1965), p. 28. The Iranian demon Nā̊ŋhaiθya is twice cited in the Avesta (Vīdēvdāt 10.9 and 19.43).
[ back ] 23. The Petersburger Wörterbuch of Böhtlingk and Roth, which cites all three interpretations, does not mention the source of this one.
[ back ] 24. See Herman Lommel, Festschrift Walther Schubring (Hamburg, 1951), p. 29, who points out that na- instead of an- in such a compound would be highly unusual, and who argues that the name “not untrue” (or “not unreal”) would be too abstract and undistinctive to be old.
[ back ] 25. This derivation is given by Yāska (Nirukta 6.13) and defended by Lommel (n. 24), pp. 29–31. The story on which the derivation is based is found in the Bṛhaddevatā 6.162–167.7 to RV 10.17.1–2: the twins’ mother Saraṇyū conceived them when she and her husband Vivasvat, in the form of horses, attempted to mate; the seed of Vivasvat fell on the ground and Saraṇyū breathed it in through her nostrils. This myth, which has a close parallel in the Purāṇas (see Lommel, p. 30), cannot have to do with the original meaning of the name Nā́satyā. The decisive objection, which will be argued later in this section, is that Nā́satyā was originally singular and designated one twin in opposition to the other. The interpretation “nose-born,” which must concern both twins equally, is thus to be rejected.
[ back ] 26. H. Brunnhofer, Von Aral bis zur Ganga, p. 99—unavailable to me but cited by Lommel (n. 24), p. 29, and H. Güntert, Der arische Weltkönig und Heiland (Halle, 1923), p. 259.
[ back ] 27. H. Güntert (n. 26), p. 259. Güntert’s derivation has been accepted by Wackernagel-Debrunner, Altindische Grammatik, vol. 2, pt. 2 (Göttingen, 1954), p. 939. M. Mayrhofer (n. 15), s.v. Nā́satyā, regards a derivation from nes- as likely and credits Güntert with the best solution.
[ back ] 28. G. Nagy (chap. 4, n. 55), p. 43 n. 121. The Old Persian form is a Median borrowing; for the verb in question, cf. Old Persian xšay-, “rule,” Avestan xšāy-, “have power,” and Sanskrit kṣay-, “possess.”
[ back ] 29. I quote from Nagy’s discussion, which also cites Hittite evidence:

For a parallel syntagma in Celtic, cf. Old Irish 3rd plural relative bertae “they who bear” < *bheronti- i̪o: likewise Gaulish dugiiontiio “they who serve,” discussed by C. Watkins, “Preliminaries to a Historical and Comparative Analysis of the Syntax of the Old Irish Verb,” Celtica 6 (1962) 24. Such a syntactical order is well-attested in Indo-Iranian: cf. Rig-Veda 1.70.5: dā́śad yó asmāi “he who awaits him,” as discussed again by Watkins, op. cit. 29 … . I add here some possible parallels suggested to me by C. Watkins:

Lūcetius, the name of one of the followers of Turnus: Vergil, Aeneid IX 590. Servius ad loc.: … lingua Osca Lucetius est Iuppiter dictus a luce. Cf. also Gaulish Leucetios, epithet of the god of war. For references and further instances (including a possible occurrence in the Carmen Saliare), cf. J. Whatmough, The Prae-Italic Dialects of Italy II, 197.

Δουκέτιος, the name of a king of the Sicels: Diodorus Siculus 11.78.7. For references and further instances, cf. again Whatmough, PID II 452.

Hence *leuketi-i̪o “he who shines” and *deuketi-i̪o “he who leads,” both nominalized. There is a parallel syntagma in Hittite: e.g. in Laws I 25, paprizzi kuiš “he who defiles” (a well, in this case); also, in an Akkadian-Hittite vocabulary (Keilschrifttexte aus Boghazköi I 42 31), the Akkadian participle ḫābilu “gewalttätig” is glossed as dammešḫiškizzi kuiš, literally “welcher schädigt.”
[ back ] 30. The fact that a god is the subject of the activized verb in this reconstruction of the divine name Nā́satyā is to be noted; cf. chap. 4, n. 58 above.
[ back ] 31. The epithet párijman, “traversing,” is also used of the twins as a pair (e.g., RV 1.46.14). For the dual form Nā́satyā as originally meaning “Nā́satya- and his brother,” cf. J. Wackernagel, Zeitschrift für vergleichende Sprachforschung 23 (1877): 302 ff., on the Homeric dual Aíante as originally meaning “Ajax and his brother.”
[ back ] 32. Cf. also RV 1.117.4, another reference to the myth of Rebha, which says: “you two, by your miraculous powers, put him back together again when he had come apart” (sáṃ táṃ riṇītho víprutaṃ dáṃsobhir). For other references to this myth, see K. F. Geldner, Der Rig-Veda, vols. 1–3 [Harvard Oriental Studies 33–35] (Cambridge, 1951), on RV 1.116.24.
[ back ] 33. Geldner takes pitṛ́bhya ā́ as “to his parents,” but the Rig-Veda always uses the dual when the meaning is “parents.” See Geldner on RV 1.116.3–5 for other references to the myth of Bhujyu.
[ back ] 34. RV 1.117.24 and 10.65.12.
[ back ] 35. Vedic nírṛti- can be either the goddess or the concept of “destruction” or “dissolution,” and it is not in fact clear which is meant in this verse. For a comparison of the goddess with the Roman Lua Mater, see G. Dumézil, Déesses latines et mythes védiques (Brussels, 1956), pp.107–115. Related to nírṛti- is the verbal adjective nírṛta-, which is applied to the same Vandana in 1.119.7: “Vandana, who was decomposed by old age (nírṛtaṃ jaraṇyáyā), you miracle-workers put back together again (sám invathaḥ) as craftsmen do a chariot.”
[ back ] 36. For more on the relationship between the twins and the Dawn, see n. 66 below; for general remarks on this relationship see Donald Ward, The Divine Twins (Berkeley and Los Angeles, 1968), pp. 11 and 15.
[ back ] 37. RV 1.183.6, 1.184.6, and 7.73.1; the same phrase occurs in RV 1.92.6 to Uṣas.
[ back ] 38. They are called “husbands” (pátī) in RV 4.43.6, “wooers” (varā́) in RV 10.85.8–9; Sūryā́ mounts their chariot in, e.g., RV 5.73.5 (ā́ … vāṃ sūryā́ ráthaṃ tiṣṭad).
[ back ] 39. The duhitā́ sū́ryasya mounts the chariot of the twins in RV 1.116.17, 1.118.5, and 6.63.5; she “chooses” their chariot in RV 1.117.13 and 4.43.2.
[ back ] 40. The Latvian Dieva dēli are actually called their sister’s suitors; see Ward (n. 36), pp. 10–11, who also discusses the Greek Dioscuri and their sister Helen; cf. also Gregory Nagy, “Phaethon, Sappho’s Phaon, and the White Rock of Leukas,” Harvard Studies in Classical Philology 77 (1973): 162–177.
[ back ] 41. For the cult title and literary epithet of the Dioscuri, see A. Furtwängler, “Dioskuren,” in Roscher’s Lexikon (chap. 3, n. 12), vol. 1, no. 1, cols, 1163–1164. D. Ward (n. 36), pp. 14–15, 18, summarizes the evidence for the Indo-European origins of the twin’s functions as “saviors” and “healers.”
[ back ] 42. The same episode was narrated in the Cypria; fragment 11 (Allen) tells how Lynceus spied the Dioscuri before the battle began. Apollodorus 3.11.2 also narrates the episode.
[ back ] 43. In the actual narrative Pindar says that when Polydeuces reached Castor, the latter “was not yet dead, but still gasping” (l. 74): καί νιν οὔπω τεθναότ’, ἄσθματι δὲ φρίσσοντα πνοὰς ἔκιχεν. Apollodorus 3.11.2, however, says that Idas “killed” (κτείνει) Castor, and that Polydeuces refused immortality from Zeus “while Castor was a corpse” (ὄντος νεκροῦ Κάστορος). Pindar’s word φθιμένου, “perished,” expresses the underlying reality of the situation, which is that Castor, as a “mortal,” had to “die.” Fragment 6 of the Cypria (Allen) expresses this reality in terms of Castor’s “fate”: Κάστωρ μὲν θνητός, θανάτου δέ οἱ αἶσα πέπρωται, / αὐτὰρ ὅ γ’ ἀθάνατος Πολυδεύκης, ὄζος Ἄρηος, "Castor, on the one hand, was mortal, and a fate of death was allotted to him, but Polydeuces, scion of Ares, was immortal." As a combination of immortal and mortal elements, the Dioscuri, even as a pair, experience a regular alternation between “life” and “death.” Such an alternation is precisely what characterizes the twins in our earliest evidence, Odyssey 11.302–304: οἳ καὶ νέρθεν γῆς τιμὴν πρὸς Ζηνὸς ἔχοντες / ἄλλοτε μὲν ζώουσ’ ἑτερήμεροι, ἄλλοτε δ’ αὖτε / τεθνᾶσιν, "Obtaining honor from Zeus even beneath the earth, now they live, on alternate days, and now they are dead." Iliad 3.243–244 is not inconsistent with this passage, although it is briefer, and Pindar Nemean 10.55–57 follows a similar tradition.
[ back ] 44. Cf. Zeus’s speech to Polydeuces in Nemean 10.80–82: Ἐσσί μοι υἱός· τόνδε δ’ ἔπειτα πόσις / σπέρμα θνατὸν ματρὶ τεᾷ πελάσαις / στάξεν ἥρως, "You are my son; but as for this one, a hero afterwards let drip his mortal seed when he approached your mother as her husband."
[ back ] 45. Cf. D. Ward (n. 36), pp. 4–5 and 12–14, who emphasizes that dual paternity, frequently involving an opposition between an immortal and a mortal father, is a common feature in twin mythology the world over.
[ back ] 46. Cf. also RV 5.76.3, which bids the twins to come “day and night” (dívā náktam), but apparently a third time as well—namely, “midday” (madhyáṃdine).
[ back ] 47. It is also probable that the Vedic twins were identified with the morning and evening stars; cf. D. Ward (n. 36), pp. 15–18, who cites bibliography and also considers the comparative evidence of other Indo-European twins; cf. also Gregory Nagy, “Phaethon, Sappho’s Phaon, and the White Rock of Leukas” (n. 40), pp.172–173, n. 94.
[ back ] 48. S. Wikander, “Nakula et Sahadeva,” Orientalia Suecana 6 (1957): 66–69.
[ back ] 49. This is just part of a larger argument that Wikander put forth in “Pāṇḍavasagan och Mahābhāratas mytiska förutsättningar,” Religion och Bibel, Nathan Söderblom-sällskapets Årsbok 6 (1947): 27–39 (translated into French by Georges Dumézil in his Jupiter Mars Quirinus [Paris, 1948], 4: 37–53). In this earlier study, Wikander dealt with the structure of the Pāṇḍavas, the five heroes of the Mahābhārata, as a group. All five are the sons of different gods, and Wikander convincingly showed that the fathers represent the old trifunctional scheme which Dumézil has discovered and elaborated. Wikander showed further that the five sons preserve archaic representations of their fathers, and of the three functions with which they are associated. In “Nakula et Sahadeva” (n. 48), Wikander went on to study separately the representatives of the third function, the twins Nakula and Sahadeva, as they preserve old oppositions that are not directly attested for their fathers. Dumézil has summarized and expanded all of Wikander’s findings in the first part of his Mythe et épopée, vol. 1 (Paris, 1968), which is now the best source to consult; for the structure of the five Pāṇḍavas, see pp. 53–102, and for the twins in particular, pp. 73–89.
[ back ] 50. Wikander, “Nakula et Sahadeva” (n. 48), pp. 72–73. The author also shows here that Sahadeva is differently characterized as “modest, correct, obedient,” and the like. D. Ward, “The Separate Functions of the Indo-European Divine Twins,” in J. Puhvel, ed., Myth and Law among the Indo-Europeans (Berkeley, Calif., 1970), pp. 193–202, has shown that this feature of Sahadeva’s character is paralleled in other Indo-European twins, and is thus highly traditional. I cannot, however, agree with Ward’s main thesis that the “modest” Sahadeva (and the Indo-European twin he represents) is differentially associated with Dumézil’s third function. Wikander (n. 48), pp. 75–76, in showing that Sahadeva is associated with Yudhiṣṭhira while Nakula is assocated with Bhīma in battles of the Mahābhārata, demonstrated that Sahadeva is differentially associated with the first function, and Nakula with the second. Dumézil, Mythe et épopée (n. 49), 1: 81–86 devotes a separate section to “Sahadeva et la première fonction”: he calls Sahadeva an “auxiliaire de la première fonction” on p. 86. As we shall see below, Sahadeva is to be connected with the son of Dyaus, and one would naturally expect this twin to be ranked above the son of Sumakha.
[ back ] 51. Wikander (n. 48), pp. 71–72.
[ back ] 52. 1.139.5533, 5.47.1832, 5.49.1996, 5.89.3168, 8.76.3814, 9.10.477.
[ back ] 53. Wikander, pp. 73–74.
[ back ] 54. Draupadī, Bhīma, and Arjuna also die in this scene, and Yudhiṣṭhira also interprets their deaths as the result of peculiar moral flaws; see Dumézil, Mythe et épopée (n. 49), 1: 81–82, who shows that essential features of the characters in question have in this scene been turned into flaws.
[ back ] 55. Sahadeva’s “intelligence” is also a marked feature in two aberrant traditions of the Mahābhārata, as Dumézil (n. 49), pp. 82–85 has shown. The first of these is the Persian account of the “Sons of Pan” (see Dumézil, pp. 82–83). According to this, each of the five sons received a particular talent as a result of his teacher’s prayers; Sahadeva, “who looked for wisdom and who did not speak unless spoken to, asked for the science of the stars and a knowledge of hidden matters.” The second aberrant version is that of the eighteenth-century Swiss Colonel de Polier, who studied with an Indian teacher and whose extensive notes, including a résumé of the Mahābhārata, were published as La Mythologie des Indous by his cousin, the Chanoinesse de Polier, in 1809 (see Dumézil, pp. 42–43, for the nature and value of this text). In the Mahābhārata of the Colonel de Polier, the five Pāṇḍavas are all characterized at their births, and Sahadeva is called “the most enlightened of mortals, the most perspicacious, and the most learned in the knowledge of past, present, and future” (see Dumézil, p. 84). The fault which causes Sahadeva’s death in this version is more particular than it is in the vulgate: Sahadeva did not tell his brothers that their mother Kuntī had had a son (Karṇa) by the god Sūrya before they were born, “although he had penetrated this mystery by means of his great intelligence” (see Dumézil, p. 85).
[ back ] 56. See p. 48 in Dumézil’s translation of Wikander’s 1947 study (n. 49 above).
[ back ] 57. Wikander, “Nakula et Sahadeva” (n. 48), p. 76. Sahadeva expresses his preference for and competence in handling cattle in 4.3.67–72 and 4.10.288–293.
[ back ] 58. Dumézil, Mythe et épopée (n. 49), 1: 87–89. The closest parallel is the Iranian female pair Drvāspā, “Mistress of healthy horses,” and Gə̄uš Tašan, “Builder of the cow,” or Gə̄uš Urvan, “Soul of the cow” (see Dumézil, pp. 88–89). For the Indo-European comparison, Dumézil cites somewhat looser parallels in Scandinavian, Italic, and Greek (pp. 87–88).
[ back ] 59. Wikander, “Nakula et Sahadeva” (n. 48), p. 79.
[ back ] 60. Cf. Homeric Hymn 33.1–2, where both twins are first called “sons of Zeus” and then “sons of Tyndareus”: ἀμφὶ Διὸς κούρους ἑλικώπιδες ἔσπετε Μοῦσαι / Τυνδαρίδας
[ back ] 61. Wikander, “Nakula et Sahadeva” (n. 48), pp. 79 ff. For the evidence that led Wikander to this conclusion, see below in text.
[ back ] 62. Wikander, “Nakula et Sahadeva” (n. 48), pp. 79, 81–82.
[ back ] 63. The origin of such oppositions may have been dual dvandva compounds which split and resulted in two separate dual terms (ekaśeṣa-s); see Wikander (n. 48), p. 79.
[ back ] 64. These figures are based on my own count; Wikander (n. 48), pp. 81–82 reports only eight cases of co-occurrence for vṛ́ṣaṇā and only twenty for dasrā́.
[ back ] 65. It should be noted that this strophe occurs in a catalogue of gods whom Dumézil has identified as the Indic gods of the three functions; such catalogues are rare in the Rig-Veda, and highly traditional. Strophes 1–6 of RV 2.41 are devoted to the gods of the first two functions, and strophes 7–9 are devoted to the twins, the representatives of the third function. For Dumézil’s analysis of this hymn, see his Tarpeia (Paris, 1947), pp. 45–56, and Mythe et épopée (n. 49), 1: 51.
[ back ] 66. Earlier in this section there was a brief discussion of a mythological relationship between the Nā́satyā and the dawn goddess Uṣás; subsequent remarks on gómat and áśvāvat, the two adjectives which serve to oppose the twins to each other, permit further remarks on the relationship with Uṣas. The two adjectives in question occur simultaneously in the Rig-Veda only in references to the twins and to Uṣas: RV 1.92.14 invokes Uṣas with the vocatives gómati áśvāvati, “rich in cattle, rich in horses”; in RV 1.48.2, 1.123.12, and 7.41.7, Uṣā́sas, “Dawns,” in the nominative plural, is modified by áśvāvatīs gómatīs, “rich in horses, rich in cattle.”

What is an opposition in the case of the twins, however, cannot be so in the case of Uṣas, a solitary figure. As the “Dawn goddess,” furthermore, Uṣas has more to do with cattle than with horses. Besides her well-known “ruddy cows,” we also have the formal evidence of the adjective gómat, which twice modifies “dawns” in the plural, without any accompanying form of áśvāvat (RV 1.113.8 and 2.28.2); the adjective áśvāvat, on the other hand, is never used of Uṣas without an accompanying form of gómat. Uṣas is, in fact, characterized by her cattle; one wonders whether Rā́trī, “Night,” in the old pair Uṣás-Rā́trī (see G. Dumézil, Déesses latines et mythes védiques [Brussels, 1956]), was once somehow associated with horses. This would correlate well with Yāska’s statement (see text at n. 47 above) that Rātrī and Uṣas were the simultaneous mothers of the twins.
[ back ] 67. Wikander (n. 48) p. 79.
[ back ] 68. Wikander (n. 48), p. 79, and again on p. 81 as follows: “pour ce qui concerne les termes nāsatya et dasra, on n’observe nulle part, dans les texts vediques, une relation claire entre les deux, ni d’opposition ni d’autre sorte (as far as the terms nāsatya and dasra are concerned, nowhere in the Vedic texts does one observe a clear relation between the two, either of opposition or of another sort).” With one reservation (see n. 69 below), I agree with Wikander’s statement, but his next step is misleading. Since there is no phraseological evidence in the Rig-Veda for an opposition between Nā́satyā and dasrā́, Wikander attempts to establish a Rig-Vedic opposition between Nā́satyā and divó nápātā, and to infer an opposition between Nā́satyā and dasrā́, given the equation dasrā = divó nápātā (see earlier in this section). The problem with this argument is the supposed opposition between Nā́satyā and divó nápātā, which, like the supposed opposition between Nā́satyā and dasrā́, simply does not exist. Wikander argues its existence on the basis of RV 1.117.11–12, a two-strophe passage containing the name Nā́satyā in strophe 11 and the epithet divó nápātā in strophe 12. But the term Aśvínā is also present in each strophe, and there are thus two genuine oppositions in the passage (Nā́satyā/Aśvínā and divó nápātā/Aśvínā). Each opposition is articulated within a single strophe, as is required; the argument that terms in different strophes (Nā́satyā and divó nápātā) are opposed to each other seems to me to be a pure figment.
[ back ] 69. There are only four cases of co-occurrence, namely 1.3.3, 1.116.10, 1.116.16, and 1.183.4; in none of the four cases is there any opposition between the terms in the various contexts, and one context (in 1.116.16 the Nāsatyā … dasrā are invoked for having restored a mortal’s eyes) suggests that the term dasrā was perhaps used to gloss Nāsatyā: the full gloss would have been dasrā bhiṣajā(u), “you two miracle-working doctors” (see n. 70 below). One may argue, therefore, that 1.116.16, far from opposing the terms Nāsatyā and dasrā, which would have been untraditional, actually glosses the old name Nā́satyā with the phrase “you miracle-working doctors.”
[ back ] 70. Note the collocation dasrā bhiṣajā(u), “miracle-working physicians,” in RV 1.116.16 and 8.86.1.
[ back ] 71. Note the collocation dasra mantumaḥ, “you miracle-working wise one,” which is twice applied to the god Pūṣan: RV 1.42.5 and 6.56.4; for the Indo-Iranian origins of this collocation, see R. Schmitt, Dichtung und Dichtersprache in indogermanischer Zeit (chap. 2, n. 28), pp. 160–161, on Avestan dangrā mantū, “by wise resolution,” in Yasna 46.17.
[ back ] 72. I will simply note here a comparison between Vedic Nā́satyā and Greek Néstōr, which I intend to pursue in a separate study. The comparison also involves Nestor’s regular Homeric epithet hippóta, “the horseman,” which corresponds to the twins’ second name, Aśvínā. The double comparison between hippóta Néstōr and the Nā́satyā/Aśvínā raises the question of whether Nestor’s origins have to do with Indo-European twin mythology. Has the Greek Nestor, like the Avestan Nā̊ŋhaiθya, become separated from a twin brother? (For Nā̊ŋhaiθya, see n. 22 above and end note 4.)