One of the most physically and mentally demanding performances in the agnichayana and other soma rituals based on that model is the atirAtra or the overnight rite. The highpoint of it is the great ritual to the gods agni and the ashvin-s and the goddess uShas. The sAman songs to these deities from the rahasya-gAna section are sung at the beginning of this ritual. These songs collectively go by the name saMdhistotra indicating that they are sung at the junction between night and the daybreak, i.e., the dawn. That they are linked with the dawn presided by the goddess uShas is clear from one set of these songs known as the uShANi, which are thus deployed by the sAmavedic singers:
prastAva: prAti vo darshyAyato3 vA |
udgItha: o3 chChantI duhitA divo3 po3 mahi vR^iNute chakShuShA tamaH |
pratihAra: jyo3tIH kR^iNoti |
upadrava: i LA o3 nArI o3 vA hA vu vA |
nidhana: As ||
prastAva: jyo3to3 vA |
udgItha: o3 No3ti sUnarI udusriyAs sR^ijate3 sUriyAs sachA |
pratihAra: udyan nAkShatrAM |
upadrava: i LA o3rchA yi vA do3 vA hA vu vA |
nidhana: As |
prastAva: udyo3 vA |
udgItha: o3 kShatram archivat tave3d uSho3 vyuShi sUriyAsya cha |
pratihAra: saM bhakte3nA ga |
upadrava: i LA o3 mahi o3 vA hA vu vA |
nidhana: As |
The prastAva is sung by the prastotA, the udgItha by the udgAtA, the pratihAra by the pratihartA, the upadrava by the udgAtA, and the nidhana by all three.
The lyrics of this rahasya gAna are based on the mantra-s of the maitrAvaruNi vasiShTha:
praty u adarshy Ayaty uchChantI duhitA divaH |
apo mahi vyayati chakShase tamo jyotiSh kR^iNoti sUnarI ||
The daughter of dyaus has appeared now, shining as she nears;
The noble lady rolls up great darkness and makes light for us to see.
ud usriyAH sR^ijate sUryaH sachAM udyan nakShatram archivat |
taved uSho vyuShi sUryasya cha sam bhaktena gamemahi ||
The sun frees the red cows while he the radiant star rises high;
May we acquire our share [of wealth] when you O uShas and the sun’s light come forth! (RV 7.81.1-2)
This is followed by the recitation of the thousand mantra-s of the Ashvina shastra by the hotr^i, which is a great feat of memory for they are recalled from the R^igveda in a rather jumbled order, i.e., mantra-s from different maNDala come in different places in the recitations. This rite to the ashvin-s (evidently in its earlier form) is alluded to within the R^igveda itself by maitrAvaruNi vasiShTha. This reference not only supports its antiquity but also suggests it was performed at dawn:
ashochy agniH samidhAno asme upo adR^ishran tamasash chid antAH |
acheti ketur uShasaH purastAch Chriye divo duhitur jAyamAnaH ||
agni has shone brightly when kindled (using samidh-s) by us;
boundaries of darkness were visible.
Eastward is seen the banner of uShas,
one born to [indicate] the glory of dyaus’ daughter.
abhi vAM nUnam ashvinA suhotA stomaiH siShakti nAsatyA vivakvAn |
pUrvIbhir yAtam pathyAbhir arvAk svarvidA vasumatA rathena ||
The good hotR^i faces you, O ashvin-s,
now the eloquent chanter attends you with the recitation, O nAsatya-s!
Come by the paths you have formerly traveled by,
by the chariot that finds the light, laden with treasure. (RV 7.67.2-3)
The great Hindu patriot Lokamanya Tilak, based on the nature of the Vedic references to dawns, postulated that the composers of the veda lived in, or were familiar with, extreme northern latitudes. In particular, the case Ashvina-shastra impressed him the most. He realized that the whole long recitation of 1000 mantra-s was completed before sunrise, which implied a long dawn, much longer than the ones we see in India. Based on this and other allusions he outlined this hypothesis of his in the “Arctic Home in the Vedas.” The title and the form he presented the hypothesis was somewhat unfortunate because it gave rise to the impression that the ancestors of the Indo-Aryans lived inside the Arctic circle (i.e. 66.5°N latitude). While that is unlikely, it is clear from several lines of evidence in the R^igveda, chiefly, familiarity and fear of snow/cold winters, centrality of the fire cult, and the natural habitat of Ephedra (soma), that the ancestors of the Indo-Aryans, at some point temporally close to the composition of the RV, lived at a fairly northern latitude in Eurasia. Moreover the evidence from the “fire in water legend” places them in the vicinity of the Caspian or Black Seas, which have the characteristic geological setting with fossil hydrocarbon-powered mud volcanoes that could inspire this legend. On the other hand, linguistic evidence for contacts with Uralic place them even further north, perhaps in the zone just north of the above lakes closer to the Urals (~47-50°N latitude). This contact with Uralic is also consistent with the specific evolutionary connections between Baltic and Slavic one hand and Indo-Iranian on the other. This suggests that early Indo-Iranian was positioned in proximity to these northerly branches of IE. In light of these points, we are sympathetic to Tilak’s idea that it is likely that the Vedic references point to long dawns. Indeed, the cognate goddess presiding over the dawn was an important one in several branches of Indo-European, who appears to have faded away in several branches in subsequent times. Thus, cognates of Indo-Aryan uShas include, Iranian uShA (as in the Avestan incantation uShAhin gAH; also vî-daêvô-dAta 19.28 and yasna haptanghaiti 50.10), the Greek Eos, Latin Aurora and Lithuanian Aushrine. Aushrine seems to have been conflated with Venus in Lithuanian legend after the destruction of their ancestral religion (cf. indraja becoming Lithuanian name of Jupiter). All these cognate deities are described using poetic phrases similar to those used for the uShas in the veda. Thus, it appears that dawns were a prominent phenomenon in the Indo-European homeland, just as implied by the Vedic memory of the same.
Returning to Tilak’s interpretation of the Vedic dawn, there is one mantra in the RV relevant to his hypothesis, which has been difficult to understand. Tilak himself took it as evidence in support of his contention while other interpreters of the RV have tended to look at it differently. We revisit this mantra and see what it might mean.
shashvat puroShA vy uvAsa devy atho adyedaM vy Avo maghonI |
atho vy uchChAd uttarAM anu dyUn ajarAmR^itA charati svadhAbhiH ||RV 1.113.13 by kutsa A~ngirasa
shashvat= continually; purA= formerly uShAH= uShas, nominative singular; vi= verb prefix; uvAsa= had shone, perfect, third person of verb vas (class 6); devI= goddess; atha= moreover; u= emphatic particle; adya= today; idaM= this, 3rd person singular neuter; vi= verb prefix; AvaH= shone (see below for discussion) maghonI= munificent lady.
atha= now ; u= emphatic particle; vi= verb prefix; uchChAt= may light (see below); ut .atarAn= future, the terminal nasal is pronounced as pure anunAsika in recitation; anu= along; dyUn= days, masculine accusative plural; ajara= undecaying; amR^itA= immortal; charati= moves, present, 3rd person singular; svadhAbhiH= by her own powers, feminine instrumental plural.
Thus with the above analysis one may translate the mantra thus:
Formerly, the goddess uShas had continually shone forth;
now then the munificent lady shined this [light] today.
Now then, may she light up the days that will come along;
immortal and undecaying she moves on by her own powers.
Now this mantra poses the following problems of note:
1) The first pAda has been interpreted by most as meaning from the former times the goddess uShas has been continually shining forth. This is taken as a symmetric contrast to the future expectation expressed in the 3rd pAda for the days that will come along. But the point to note is that the verb vi-vas is used in the perfect tense (vi-uvAsa; pANini liT): it is a regular active perfect of the class 6 verb vas. Now traditionally liT is used to convey parokSha-bhUta or a distant past that has not been witnessed directly by the speaker. Indeed such a narrative tense is shared between Sanskrit and Greek and commonly used in both traditions while recounting AkhyAna-s or epic events. Hence, in the strictly grammatically sense the pAda would mean: “Formerly, the goddess uShas had continually shone forth” just as the great commentator sAyaNa and Tilak had interpreted it. Thus, Tilak’s idea that the mantra-composer was familiar with a former time when there was a continuous dawn is not technically incorrect. This might imply that the author of this mantra himself was not in the region with the long or the continually shining dawn but his ancestors were and thus he had heard of the same.
2) vi-AvaH is a peculiar verb form that does not appear in the regular conjugation of the class 6 verb vi-vas. It is found only in the older parts of the veda. sAyaNa informs us in different places in his commentary that it is derived from either la~N (imperfect past) or lu~N (aorist past) via luk, i.e. elision of the vikaraNa or the conjugational affix inserted between the root and the termination, followed by further Vedic-specific transformations (lopa of hal, i.e. pANini’s designation for terminal consonant). Of the past tenses, la~N is termed anadyatana-bhUta (used for a past that has not taken place today). While la~N is often used in conjunction with liT (parokSha-bhUta), in this context the adverb adya is used suggesting it is a past proximal to the present events. Hence, it should be taken as a lu~N, i.e. the aorist which can express adyatana-bhUta (past continuous with the present).
This interpretation of vi-AvaH is consistent with another occurrence of it in a mantra of dIrghatamas auchAthya in a similar context relating to the dawn Ashvina ritual:
abodhy agnir jma ud eti sUryo vy uShAsh chandrA mahy Avo archiShA |
AyukShAtAm ashvinA yAtave rathaM prAsAvId devaH savitA jagat pR^ithak || (RV 1.157.1)
The verbs herein are: abodhi= was awakened, 3rd person singular atmanepada aorist of class 1 verb budh; ud-eti= rises up, 3rd person singular present of class 2 verb i; vi-AvaH= shone forth, like above R^ik in the context of uShas; AyukShAtAm= was yoked, 3rd person dual atmanepada aorist of class 7 verb yuj, technically ayukShAtAm but the first syllable is metrically elongated; pra-asAvIt= 3rd person singular aorist of class 6 verb sU;
Thus the above mantra has 5 verbs of which 4 are indicate past actions that precede the rising of the sun indicated in present tense. Of these 3 are clear lu~N, i.e. aorists, indicating past actions just prior of the rising of the sun. Thus, the 4th of this group AvaH is also likely to follow them in being a specialized lu~N derivative occurring only in the veda. In his commentary on this mantra, sAyaNa indicates it is exactly that – a lu~N, which has been modified in Chandas.
3) vi-uchChAt occurring the 3rd pAda of RV 1.113.13 is another unusual verb form. sAyaNa tells us that it is a conjugational form of vi-vas just like AvaH and the perfect form uvAsa. He further informs us that it is a veda-specific conjugational form, leT. This form is taken to imply AshIH, i.e. the hope that something will happen (generally taken to imply a mood stronger than opative, li~N, and weaker than imperative, loT); the subjunctive mood of the white indologists.
One may ask if there are any other definitive indicators in support of Tilak’s contention that the Vedic dawns were long. We do think that the very prominence of the dawns, the mention of multiple changing colors from purple to “white” are suggestive of this. Moreover, we also think the legend of indra smashing the cart of uShas is another point supporting this. On multiple occasions indra is described as setting free or giving rise to the uShas and sUrya (e.g. “yaH sUryaM ya uShasaM jajAna” in RV 2.12.7). On the other hand he is described in 4 different sUkta-s as smashing the cart of uShas with his vajra: RV 2.15.6; RV 4.30.8-11; RV 10.73.6; RV 10.138.5. The cart of uShas is described as being slow (ajavasaH in RV 2.15.6) as compared to the fast chariot of indra. It is also explicitly described as being an anas, the cognate of Latin onus – a slow burden-bearing vehicle. This suggests that the legend arose from the dawn being drawn out as alluded to in RV 1.113.13. While the beauty of the dawn is celebrated in the deity uShas, on the flip side the long dawn delays actual day break. Hence, she is described as causing trouble (RV 4.30.8) while expanding her glory (RV 4.30.9). Thus, indra is seen as clearing the way for the sun to rise by causing uShas to flee (ending dawn) by destroying her cart with a strike of his bolt. Hence, this legend, which was seen as a mysterious one by the white indologist Oldenburg, is in our opinion a fairly clear memory of long dawns in the R^igveda.
In conclusion, while it may be a far stretch to place the home of the Indo-Aryans inside the Arctic circle, we do feel Tilak’s observation regarding the long dawns in the veda is valid and consistent with the idea of the ancient home of the Indo-Aryans north of the Caucasus.