The Sarasvatī in the RV (Ṛgveda) is a very large river fed by many tributaries. Along its banks thrived, says the RV, the 5 Aryan tribes, chiefly the Purus. The river dried up c1900 BCE. The name is cognate with Iranian Harxvaiti and its first element has cognations in other IE (Indo-European) tongues. It is thus one of the keys for dating the RV, and much of ancient Indian history and for settling many aspects of the IE scene.
1. The name Sarasvatī means ‘she who has sáras’. It is feminine and it is formed with the noun sáras and the taddhita matup suffix (Pāṇini 8.2.9, 4.2.72 etc) –vat > fem -vatī which indicates possessing (as bhagá-vat ‘possessing wisdom, prosperity, love’, mánas-vat ‘possessing spirit’).
The noun sáras denotes ‘lake, pond, pool’! It is cognate with Iranian *Harax- (*harah-) which is isolated in that language and does not yield a meaning, as we shall see below (§4ff). It is also cognate with Greek helos ‘swamp’. In this respect Sarasvatī is ‘she who has lakes’. But how can a river possess lakes? Well, this could refer to the pools formed at the delta of the river as it flowed into the ocean (or disappeared in the Batnair desert). Or it could refer to a large lake or lakes on the Himalayas from which it flowed down into the valley.
However, sáras means also any ‘moving/flowing sheet of water’. As such the name Sarasvatī means ‘she who has currents, eddies, swirls, turbulences, whirlpools’ in its streaming water. And this, would be a much more reasonable interpretation since it denotes a continuous aspect of the river, rather than the lakes/pools which indicate an initial and/or terminal aspect only.
Moreover, the noun sáras comes from dhātu √sṛ > sarati/sisarti (1st and 3rd class respectively) which in the sense gatau means ‘flowing, moving, rushing’. From this dhātu (=lexical seedform, root) we have the full conjugation of the two verbs: sarati he/she/it flows, sasrāva has flowed, sariṣyati will flow, and so on; sisratuḥ ‘the two flow’ etc. Also, many nouns and adjectives: sṛta flowed; sṛti flow; sara flow/flowing; sarit ‘brook, stream’; sāra ‘pith, essence’; etc, etc.
Then, this verb sarāmi ‘I flow/rush’ is cognate with Greek hallomai Latin salio (<salire) and Tocharian B salate – all meaning ‘leap, rush’. This strengthens the meaning of movement and rushing rather than the immobile lake/pond/pool, which must be, like Greek helos, a later meaning.
All this may mean very little at present but the significance will emerge below, when I discuss the relation with Iranian Haraxvaiti (§ 4-5).
2. The river Sarasvatī as presented in the hymns of the RV gives us a key to the date of the composition of the samhitā.
The river is mentioned in all the books of the RV except the 4th. The name denotes also a goddess and 2.41.16 calls Sarasvatī ‘best river nadītamā, best mother ambitamā and best goddess devitamā’. Elsewhere she is a celestial river and, of course, a goddess (of plenty and wisdom). In this paper I deal with the river only. She is a mother because she nourishes the Aryan tribes (in 6.61, 10.64, 10.177) providing sustenance and prosperity. 6.52.6 says that she is fed pinvamānā by three or more (lesser) rivers, while Yajurveda 3.4.11 states that she is augmented by 5 tributaries.
An important point is that RV 7.95.2, a very early hymn (Books 3, 6 and 7 are generally regarded as the earliest), says the river flows “pure from the mountains to the ocean” girī́bhya ā́ sumudrā́t. Then 6.61. 8-13 lauds the river as endless, swift-moving, roaring, most dear among the sister-currents and, together with her divine aspect, nourishing the five tribes. Several other verses in the Rigvedic hymns (7.96; 8.21; 10.64 and 177) praise the river for its greatness and pray that it continues to give sustenance to the peoples along its banks.
A simple fact should be borne in mind: no other river receives such praise, or, indeed, any praise in the hymns. There is no hint anywhere in the RV, no complaint or bewailing that the Sarasvatī is a desiccated river vanishing in the middle of the desert – as Manusmṛti 2.21 and scholiasts thereon say. In historic times the river is indeed shrunk and called Chaggar/Hakra or Sarsuti (<sarasvatī).
When did this happen?
I do not know. It is for archaeologists and geologists to give us the correct answer. For, if the Rigvedic references to the mighty river are a fiction then nothing in the RV can be trusted.
But reputed archaeologists, G. Possehl (1998) and B. B. Lal (2002), say that the river flowed down to the ocean before 3800 or thereabouts. Other scientists have traced the full course with satellite photographs (Sharma 2006). Others again say that the river dried up c1900 BCE due to tectonic adjustments, shifts of river-courses and other climatic conditions (Rao 1991, Allchins 1997). Due to the subsequent desiccation of the region, the inhabitants moved eastward to the Gangetic plain.
These considerations bear mightly on the date of the composition of these hymns, the RV and, by extension, of the presence of the Indoaryans in the region. (All this was discussed at great length in Kazanas 2009, ch 1, and 2015, passim. But more recent studies present contradictory data about Sarasvatī.)
3. The mainstream doctrine has been teaching the AIT (=Aryan Invasion/Immigration Theory). It held that the Indoaryans came from Iran and settled in Saptasindhu, the Land of the Seven Rivers, in North-West India and Pakistan of today, c1700 BCE. Earlier the date was 1500; now mainstreamers have changed it up to 2000. This change is not because the RV has suddenly acquired antiquity but in order to avoid the difficulty of having foreign intruders settling in a region which (in 1900 BCE) was being desiccated and the local population was moving eastward.
I have argued fully elsewhere that this mainstream doctrine has no basis whatever in hard evidences and facts; it is concocted out of pure conjectures without even, as is claimed, any linguistic basis (Kazanas 2009 mainly for archaeological evidences; 2015 mainly for literary, linguistic and genetic ones).
Neither the RV nor any other Indian or non-Indian document refers to this alleged entry. The theory began in France and later in England in the 18th cent. as sociological speculation to explain the caste system as being a result of an Egyptian or Mesopotamian invasion and conquest, then was turned into linguistic arguments (chiefly Max Müller) and now it is repeated quite mechanically without any rational explanations (Kazanas 2015, Introduction).
The archaeologists also have failed to find any evidences of intrusion – even Western ones, even Indians who have adopted the wretched AIT (Kazanas 2009, ch 1).
Now, clearly, if the Sarasvatī was a mighty river that flowed down to the ocean only before 3800 BCE, according to the archaeologists mentioned above, then the hymns that refer to it must have been composed before or at about that date. This applies to the part of the Yajurveda 3.4.11 that mentions the river. Consequently the Indoaryans were inhabitants of that area before that date and the Rigvedic world is anterior to the mature Harappan culture of c2800-2600 BCE. This is not surprising since many items common in the mature Harappan, like bricks iṣṭakā (first used in the Yajur Veda), fixed altars, urban settled life, iconic representations etc, are not found in the RV.
Corroboration and more evidences come from a comparison of the names Sarasvatī and Iranian Haraxvaiti.
4. Iranian (or Avestan) has no cognate for Vedic √sṛ and derivatives. Its own word for lake is vairi while for water it has vār which is cognate with Vedic vār ‘water’.
The name Haraxvaiti appears in the first chapter of the Avestan Videvdād along with place-names Haetumant (=Helmand), Māuru or Margu (=Margiana), Baχδī or Baxdhri (=Bactria) and others, and, of course Haptahǝndu which transliterates Sanskrit saptasindhu (short for the Vedic sapta-sindhavaḥ in the RV) ‘seven rivers’. These are places, we are told, which the Iranians passed through before they settled in South Iran and from there moved northwestward. So they record their movement out of India! Yet the AIT doctrine insists that it is the Indians who moved out of Iran into Saptasindhu despite total absence of any evidence for this!
This name too has the possessive suffix –vaiti and should mean ‘she who has harah’. But *harah- or *harax- is a stem totally isolated in Iranian: it has no other related lexemes. In this name we find yet another piece of overwhelming evidence that it is the Iranians who moved away from Saptasindhu, Bactria and so on into Iran.
The AIT in all its folly claims that the Indians and Iranians lived together in Iran in the so-called common Indo-Iranian period and spoke one language (having come there from the Russian Steppes at an earlier period). The Indoaryans then moved and at c1750, or 2000 as is the new date, settled in Saptasindhu, only to move 100 or so years later eastward and southward. In moving to their new habitat, they brought with them the memory of the river they knew in Iran and gave it to a river in Saptasindhu. But the comparison of the two names Sarasvatī and Haraxvaiti tells a different tale.
5. The AIT doctrine is attended by many absurdities.
If the Sarasvatī was a small river in the process of drying up, then the immigrant Indoaryans would have given the name to Sindhu which was a large river flowing from the mountains to the ocean. Furthermore, they would have praised that river and not Sarasvatī as endless, mighty and nourishing the tribes. But we know from hymn 10.75 that the Vedics knew of both Sarasvatī (and place it in its correct geographical place among the rivers) and Sindhu, since both are mentioned. Moreover, hundreds of sites along its banks have been unearthed by archaeologists.
Furthermore, the order of the rivers as given in hymn 10.75 is from east to west. If the Indoaryans had come from the west then their knowledge of the rivers would have been in the reverse order – from west to east. It may be argued that many folks tend to enumerate locations from east to west because of the (worship of the) sun, but if the Indoaryans had just arrived there then moved eastward, they would have not stayed there long enough to get accustomed to the geography and to compose those hymns. The hymns suggest they had been at Saptasindhu a very long time.
But the most glaring absurdity is in the disregard of the names. No adherent of the AIT ever discusses the two names.
In the entire extant Avestan literature and its language the word *haraḥ- remains an isolated miserable orphan: it has not one cognate. But according to the laws of change and correspondence regarding Old Iranian (Avestan) and Old Indic (Vedic), *harah- is the exact equivalent of Vedic saras. The correspondences are many: asura ahura, dasyu ‘demon’ dahyu/daŋhu, nas/nāsā ‘nose’ nāh/nāŋhā, soma ‘the drink’ haoma. Vedic |s| is Avestan |h|.
The significance here is that |s| is declared by all indoeurepeanists to be original and |h| to be a later change. So, while the Iranians stayed in the alleged original habitat, they lost the original IE sound |s|. On the contrary, the Indoaryans moved far, yet managed to retain the original |s|. Thus Old Indic is closer to Proto-Indo-European. This is totally unreasonable in the context of the AIT.
Moreover, and this is even more surprising. The sedentary Iranians lost the dhātu √sṛ and all its derivatives except the one *harah- but the mobile, nomadic Indoaryans retained this. For it would be utterly ludicrous to suppose that they came to Saptasindhu and here developed (from *harah-) the complex lexical family of sṛ/sar which are cognate with other IE words in Greek etc (see above, §1).
The Avestan Haptahǝndu is also revealing: this too is a transliteration of Sanskrit Saptasindhu. But the word ‘river’ in Avestan is θraotah– (Sanskrit srotas), ravan and rauδah-. The word -hǝndu is met only in much later Iranian in the sense ‘Hindu/Indian’! So here again the Iranians carried a memory of Saptasindhu in their new habitat.
In ch 4 of my Vedic and Indo-european Studies (2015) I discuss very fully the Vedic-Avestan relations and correspondences and show that Avestan is far more recent (later) than Vedic. Vedic itself is a derivative of the Proto-Indo-European language but far closer to it than any other extant IE tongue.
6. Thus the Sarasvatī is one of the keys for unlocking the puzzle of historical dates and the origin of the Indoaryans. And it has two aspects: the flow of the river itself, which is a matter for archaeological and geological investigation, and the name and its Avestan correspondence which is a matter for linguistic investigation.
In this paper I have provided the linguistic evidence which shows that the Indoaryans retain faithful memory of the original Proto-Indo-European sounds and roots and derivatives. The Iranians on the other hand lost all this. In fact the name Haraxvaiti is a straight transliteration of Sarasvatī showing that the Iranians moved away from Saptasindhu (the Avestan haptahǝndu in their Scriptures which is another transliteration of the Sanskrit) and took with them the memory of the Land of Seven Rivers as well as that of the mighty river on the bank of which they had lived.
The Indoaryans are therefore indigenous on this account.
It is up to the archaeologists and geologists now to reinforce with fresh evidence (or cast doubt on) the linguistic finds.
The Sarasvatī is only one key. There are others, as I have shown, but they belong to a different context and discussion.
Allchin B&R: 1997 Origins of a Civilisation N Delhi, Viking Penguin.
Kazanas N : 2009 Indo-Aryan Origins… N. Delhi, Aditya Prakashan.
2015 Vedic and Indo-European Studies N Delhi, Aditya Prakashan.
Possehl G : 1998 “Did the Sarasvati ever flow to the sea?” in G. Phillips et al (eds) Arabia and its neighbours (in honour of B. de Cardi), Turnhout, Bepels (335-354).
2003 The Indus Civilization Roman & Littlefield (non NBN); Delhi, Vistaar.
Rao S.R.: 1991 Dawn & Devolution of the Indus Civilisation, Delhi, Aditya Prakashan.
Sharma JR et al : 2006 ‘Course of Vedic river Saraswati…’ in Puratattva vol 36 (187-195).