An Exegesis of Ṛgvedic Chronology

Vedic literature begins with the Ṛgveda and its ten maṇḍalas, or books. Each maṇḍala is composed of hymns or sūktas, which in turn are made from verses or mantras. There are a total of 1028 sūktas and 10,552 mantras, and this table shows the breakdown:

A body of scholarly work has been dedicated to finding a chronology to the Ṛgveda. What do we mean by chronology for a text that was transmitted orally for many generations? Several Vedic hymns contain astronomical data going into the Paleolithic, so are we to say that the Ṛgveda was composed during the Last Glacial Maximum? Add to this Hindu tradition that considers the Vedas apauruṣeya, or authorless and eternal. The claim needs elaboration, for it seems to defy any historical analysis. In his book Hindu Dharma, Swami Chandrasekharendra Sarasvati takes the example of the Law of Gravitation, or the discovery of America by Columbus. In the former case, gravitation exists eternal- independent of Newton’s discovery of it. In the latter case, Columbus’ discovery did not create America, for the land already existed. This is precisely the apauruṣeya nature attributed to Vedic sūktas. Vedic ṛṣis did not create or compose them, they discovered or perceived them. This perception was called a kind of inner sight, and thus were Vedic ṛṣis called dārṣṭas, or seers. Swami Sarasvati makes clear that Vedic ṛṣis were not mantra-kārtas (mantra-creators) but mantra-dārṣṭas (mantra seers).

For a historical objective, this makes our task both easier and less controversial. Gravitation may be eternal, but Isaac Newton was a specific Englishman who lived in the 17th-18th centuries AD, and ‘revealed’ gravitation in the lexicon of his time. Similarly, while the genesis of continental America goes into the geological past, Christopher Columbus came upon it in 1492 AD and ‘made it known’ to his world. With the wisdom only a supremely ancient civilisation can have, Indian tradition does not care for such queries into Vedic literature. Would the modern world care, if the discovery of America was recorded in our annals a hundred times over, by a hundred different men? Would the names of each of their ships, or the European monarchs who financed each, have been remembered even? And if yes, how many Christophers, Johns and Ferdinands would we find in the lists? This is why multiple Viśvāmitras or Bharadvājas do not perplex Indian tradition.

Indian tradition differs in another, more profound way. Strictly speaking, it would not even care for who Columbus really was, or who financed him and to what purpose. Instead, it would wonder whether he was a Rākṣasa or a Deva, a malevolence or an upholder of dharma. It would ask what ought to be when encountering a new land and a new people, and historical Columbus would be important only for lessons on what to do or not do. In this example the native testimony stands unambiguous- Columbus was Rākṣasa. In telling his story, the Purāṇas might as well have added an avatāra to vanquish Columbusasura. A millennia or two after that, ‘modern’ historians would have ignored this mythological mess and split hairs trying to identify whether the story happened in 1492 AD, 1092 AD, or was completely fictional!

What we can aspire for though is to discern between the few relevant Viśvāmitras, or identify the specific Bharadvāja of the Ṛgveda. When bundled with testimony from linguistic, archaeological and other analysis, we need not be held back by the eternal nature of Vedic content in assigning a date to its composition and assembly.

Even as mantra-dārṣṭas, the Vedic seers perceived their sūktas at a specific point in time, and both the Ṛgveda and the Purāṇas are valid testimony in identifying these chronologies. Certainly the body of knowledge suggested in Vedic literature did not originate in a space of just 500 years, and Vedic composers built atop knowledge systems with millennia-long lineages. But the science of sound engineering, the practice of embedding into metre and hymn sacred knowledge, and the usage of these hymns in fire-based rituals- these are cultural traits that would have arisen at some point, and gained formal structure under a select group of leaders/founders.

By 3500–3000 BC India, we find evidence of fire-altars at sites like Bhirrana and Kalibangan, and increasing contact with the north-west brings true horses within the Indian geographical sphere. If we are to map this period with Vedic names, Paurāṇika genealogy suggests this to be the era of Jamadagni Bhārgava and Viśvāmitra Kauśika. A few generations after them emerge the Ṛgvedic Āṅgirasas, and true Vedic period begins. The early period is characterised by the dominance of a few ṛṣi families and primary ṛṣis, but how do we determine an ‘early’ period?

There is broad consensus that the books 2–7 form the Ṛgveda’s oldest core, while books 9 and 10 are chronologically the last. The consensus is specifically on “composition and assembly,” meaning that it bases periodisation along the timeline of metrical hymnology (composition) and formal structure (assembly), where the composition has long antecedents and the assembly forms the tape-recording- a freezing in time. Book 10 is considered to be of a strikingly different language than of the previous books, leading us to conclude that it represents a somewhat later period than the rest of the Ṛgveda. Book 9 is called the Soma Pavamāna, a kind of appendix comprising hymns dedicated to soma, and textual examination suggests that this maṇḍala was put together when the previous eight were already in place. Books 2–7 are called the family maṇḍalas, as each is dominated by hymns of a specific ṛṣi family. In his work The Rigveda — A Historical Analysis, Talageri uses internal data to periodise the family maṇḍalas. His analysis can be summed such:

Talageri’s analysis is criticised for implying a discrete periodicity between the family books, but this assertion is not necessary to grasp the broader chronology. Take the case of books 6 and 4, both by Āṅgirasa ṛṣis. We know that Gautama ṛṣis were descendants of Gotama and Dīrghatamāsa, of which the latter was contemporaneous to Bṛhaspati. A name such as Vāmadeva Gautama necessarily comes a few generations after Bharadvāja Bārhaspatya. This allows us to conclude that book 4 was not composed before book 6, or that the latter was already in place by the time Vāmadeva began composing the former. In fact the other composers of book 6, with names such as Bhāradvāja and Bārhaspatya, are clearly Bharadvāja’s immediate descendants. Vāmadeva follows him by at least a few generations. Only going by genealogy is no solution of course. Book 10 contains composers such as Māndhātṛ and Pṛthu, but it’s taken as the latest book because it was assembled last. Whatever language Māndhātṛ and Pṛthu spoke in, it could not have been the exact dialect of 3000 BC or later. But their hymns and knowledge were codified by a later culture into its own dialect. Some support for this comes from RV 8–40–12, where Nābhāka Kāṇva says that Māndhātṛ’s ancient songs to Indra and Agni are being sung ‘in a newer speech’ or navīyo avāci.

Similar conclusions can be arrived at when we map names of rulers in the Ṛgveda with their positions in Bhārata genealogy. Book 6 features Divodāsa, books 3 and 7 feature Sudās, and book 4 names Sahadeva and Somaka. In genealogy, Divodāsa precedes Sudās by a few generations, while Sahadeva and Somaka are Sudās’ descendants. Talageri’s other conclusions are supported by several linguistic scholars. For example, Edward Hopkins concluded that book 5 shows “special rapport” with book 8 while showing “no correspondence” with other family maṇḍalas. This allows us to place it later in chronology to the books 2–4, 6 and 7. At no point is discrete periodicity suggested, nor is it necessary to assume no overlap. Instead, given that we talk of composition and assembly, what is meant is that the sūktas of maṇḍalas 2–4, 6 and 7 were largely in place before those of maṇḍala 5 emerged- and further that the latter was assembled closer in time to assembly of maṇḍala 8. Talageri also quotes Theodore Proferes- “We need not rely exclusively on the anukramaṇī to affirm that there were important interactions between the priestly groups represented in Books 1, 5 and 8. As Oldenberg has shown, evidence from the hymns themselves supports this conclusion.”

We therefore have a three-fold division of the Ṛgvedic period, where in some cases there are degrees of separation between composers of different generations, while in others there are correspondences between related priestly groups. These are broad conclusions, for the Ṛgveda we know would take final shape only later under Jātūkarṇa and his disciple Veda Vyāsa:

• Early Period: Books 2–4, 6 and 7

• Middle Period: Books 5 and parts of 1, 8

• Late Period: Books 1 and 8–10

Together these books cover the entire period of the Bhārata dynasty. The earliest book, maṇḍala 6, is authored by Bharadvāja and his immediate descendants, who were contemporaneous to Divodāsa. By the final books 9 and 10, we see the appearance of Devāpi and Śāntanu, who are forefathers to the Kuru-Bhāratas. It’s not without reason that the Bhāratas are, in Talageri’s words, the people of the book as far as the Ṛgveda is concerned. Chronology of the rest of Vedic literature maps remarkably well to genealogy. The Yajur Veda makes no mention of any ruler later than Dhṛtarāṣṭra, and the Atharva Veda names Parīkṣit. Evidence suggests that the Ṛg, Yajur and Sāma Veda had some kind of distinct form by the time of Hiraṇyanābha of Kosala in 2200 BC, some seventeen generations after Rāma. Two centuries after that in ~2000 BC, under Brahmadatta of the Kurus, two ministers by the names Kaṇḍarīka and Subālaka (or Gālava) combined the texts into a whole. Gālava is credited with instituting śikṣā, or formal instruction of the Veda and kramapāṭha, a kind of recitation method. He is also a Vaiśvāmitra, so the school is deeply embedded in Vedic tradition.

But the compiler supreme, the founding father of Vedic literature, was Veda Vyāsa who lived during the Mahābhārata war. Vyāsa is credited not only with dividing the Veda into four formal texts, but also with compiling the Ṛgveda into its ten maṇḍalas, with the particular sequence of hymns. According to tradition he instructed Paila in the Ṛgveda, Vaiśampāyana in the Yajur, Jaimini in the Sāma and Sumantu in the Atharva. Of these the Atharva is somewhat different to the other three, and during Hiraṇyanābha or Brahmadatta’s era we find mention only of three Vedas. The Atharva is stamp not only of Vyāsa’s authority but also his intent. It’s a collection of rituals, formulae and procedures for everyday life- unlike the formal ritualism of the other three Vedas. It represents Vyāsa’s attempt to bring synthesis to the vast culture and religion(s) of his time, an impulse with deep roots in the Vāsiṣṭha school he descended from. The Vedas saw further divisions and branching, as Vyāsa’s disciples divided it among their own students over the generations.

Around a century after the Mahābhārata war, the first of the Brāhmaṇa texts were compiled, beginning with Pañcaviṁśa and Taittirīya and with Śatapatha and Jaiminīya in the later period. They are supplemented by Āraṇyakas and Upaniṣads to complete the Vedic corpus, all of which makes no mention of any ruler later than Parīkṣit. PC Sengupta takes a reading of astronomical data to make the following broad conclusions about composition and assembly:

• Ṛgvedic period: 4000–2000 BC

• Brāhmaṇas and Āraṇyakas: 2000–1000 BC

• Siddhāntas and Purāṇas: 1000 BC-500 AD

This is happily in line with our chronology, where pre-Vedic ancestors such as Yayāti and Viśvāmitra are placed between 4000–3000 BC, and the people of the book emerge by the beginning of the 3rd millennium BC. Knowledge and ritual have long antecedents, but under the Bhāratas is the genesis of Vedic composition and assembly, beginning with the titular founder. The problem lies in discrepancy between Ṛgvedic and Paurāṇika evidence, best considered through the cases of Viśvāmitra Gāthina and Bharadvāja Bārhaspatya.

Viśvāmitra Gāthina is primary composer of maṇḍala 3 of the Ṛgveda, along with names that appear to be his direct descendants/disciples. Analysis of this maṇḍala suggests that Viśvāmitra was contemporary to Sudās, the Bhārata king to feature in book 3. But Paurāṇika tradition tells us that Viśvāmitra lived a long time before Sudās, and precedes even Bharata. In fact, another Viśvāmitra was Bharata’s maternal grandfather. Paurāṇika evidence also tells us that Viśvāmitra’s nephew was Jamadagni Bhārgava, who features as a Ṛgvedic composer only in books 8, 9 and 10, and is absent from the entire early period. Some Purāṇas show the Kanyākubja kings to descend from Ajamīḍha in the line of Bharata, which allows us to place Viśvāmitra Gāthina as contemporaneous to Sudās. But nowhere does the Ṛgveda refer to Viśvāmitra as a Bhārata, which it easily should if his line descended from Ajamīḍha.

Confusion stems from gāthina, or from its emphasis as a patronymic for Viśvāmitra. It emerges from the root gai, which lends itself to words such as gīta, pragātha, gāyatrī and gāthā- the latter also being an Avestan word. Gāthina appears only once in the Ṛgveda, where in maṇḍala 1 a Vaiśvāmitra refers to himself as gāthina. This led to confusion in the anukramaṇīs, which took gāthina to be a patronymic. It cascaded into Paurāṇika genealogy too, where an imagined father by the name Gāthi was inserted between Kuśika and Viśvaratha. In reality, the composer of maṇḍala 3 is not Viśvāmitra Gāthina but Viśvāmitra Kauśika. Or at least the original composer was, for the Ṛgvedic Viśvāmitra goes deeper. Maṇḍala 3’s other composers are named Gāthi Kauśika, Ṛṣabha, Kata and Prajāpati Vaiśvamitras (Viśvāmitra’s sons), and evident descendants by name alone- Kuśika and Kātya. These are clearly the same people of Paurāṇika genealogies, so the question is only where in time to place them. If we follow the Purāṇas, we are forced to explain why Sudās appears in the Ṛgveda as Viśvāmitra’s patron. If we follow the Ṛgveda, we must ignore compelling synchronisms in the Purāṇas between Viśvāmitra, Jamadagni, Paraśurāma, Hariścandra and Arjuna Kārtavīrya. Epistemological validity requires we favour the Ṛgveda, since it descends to us as a preserved tape-recording, while the Purāṇas have untold discrepancies and fabrications.

But if Viśvāmitra Kauśika did indeed live during Sudās’ era, why would the Purāṇas create imagined relationships with Jamadagni Bhārgava and Aikṣvāku kings?

The theory that it was an attempt to mythify the Ṛgvedic conflict between Viśvāmitra and Vasiṣṭha doesn’t hold water. The Ṛgvedic conflict is no conflict at all, merely inferred because Sudās’ priest at an early period is Viśvāmitra, and at a later one is Vasiṣṭha. No direct evidence can be gleaned from the Ṛgveda for a great conflict between these two ṛṣis, which the Purāṇas would later (allegedly) exaggerate into countless myths. A proposed resolution to these problems is based on realising the full import of two things- 1) the tape-recording aspect of the Ṛgveda and 2) the east-to-west movement of people evident in it. To parse this, let us engage in a thought experiment.

Imagine recording today a video submission for a job interview. You record in this video the basic details of your life, upbringing, education and career so far. Additionally, you record some technical knowledge pertaining to the role you apply for. At the instance of recording, you freeze in time the specific accent, vocabulary and knowledge you possess. You may quote something your professor once taught you, but you quote him in English for this interview, while the professor had originally spoken to you in Hindi. You make an error and name 2011 as the year you passed out of college, though you passed out in 2010. During the interview, when you talk about a previous job you enjoyed, your voice carries passion and speaks with conviction. In a different segment, where speaking of your own weaknesses, you are cautious and speak with hesitation. These differences show up in your volume, vocabulary and dialect. Opting not to bother the video’s imagined viewer with pedantic details, you paint broad strokes for much of your early life. So you omit entirely that you lived in Mumbai for a couple of years, for mentioning it serves no purpose in this interview. In fact you say you’ve been living in one city, Gurgaon, for the past ten years; though you interned in Kolkata and Coimbatore for a couple of months in between.

Imagine further that this recording then becomes the only remnant of your life. Once you are gone, anyone wishing to learn about you has only this recording to extract from. Anything present in it is bound to characterise you, while anything absent will be taken as evidence of absence. Thus the future historian concludes that you were an English-speaker, not realising, or not considering it important, that you actually spoke Hindi for most of your life. The historian assumes that all you ever knew is what’s represented in this interview. All knowledge that was irrelevant to the interview’s technicalities is absent, and the historian assumes you never possessed it to begin with. The historian of course concludes you to be a native of Gurgaon, and has no inkling that you might have been born in Benaras- since you never mentioned that in the interview.

In other words, as precise as your tape recording is, as misleading a guide it can be for a historian interested in reconstructing your life. She’ll surely get some things right. You were in fact as proficient in English as a native speaker would be, so your characterisation as an English speaker will never be falsified. You lived much of your life in Gurgaon and were shaped by it, so inferences drawn from taking you to be a native of Gurgaon will often hit the bulls-eye. But equally will wrong inferences be drawn. The historian will assert, on the validity of her discipline, that your quoted professor was also an English speaker- in the process erasing a Hindi-past, however inadvertently. She might even conclude, given that you’re a native of Gurgaon to her, that your professor also lived in Gurgaon- thereby changing entirely the geography of someone.

Such are the problems with the Ṛgveda. No amount of linguistic and scholarly analysis can prove that hymns of the Ṛgveda, regardless of when they were frozen in time (assembly), were not dialectical re-expressions of previous knowledge (composition).

We already know that astronomical data, however valid, can transmit through the ages and be embedded into later prayers- leading us into confusion if we’re not careful. The same can happen with any kind of data, and the only thing actually ‘recorded’ in the Ṛgveda is geography and language (and some history by extension). We know where the Ṛgvedic people lived and what language they spoke, but we can conclude nothing about their past (prior to the ‘recording’) from the recording alone. Even the east-to-west movement is an inference, and inferring it completely in any case helps our resolution.

When we see in the Ṛgveda a movement of people from the Gangetic Plains to the Sarasvatī’s, we should ask- what moved with these people? Certainly language and dialect did, for we know the Sarasvatī-Sindhu region to be the dispersal point for PIE. But more broadly, when people move then two things can move with them- material and ideas. The movement of material is discernible in archaeology, through changes in pottery styles or burial cultures for example. The movement of ideas is far harder to discern, but it happens all the same. When we see Ṛgvedic ancestors having come from the Gaṅgā, we must conclude that some Ṛgvedic ideas came from there too. Viśvāmitra’s own enlightenment occurred along the Kosi river (by some accounts), and the early Bṛhaspatis and Bharadvājas were rooted in Vaiśālī. Even the Vāsiṣṭhas are placed in Ayodhyā before anywhere else, and only the Bhārgavas among the primary families have origins all over India- but they are not the primary Ṛgvedic ṛṣis.

This explains Viśvāmitra Kauśika’s coexistence with Sudās in the Ṛgveda. The gāthina that Sudās patronised did indeed ‘compose’ and ‘assemble’ maṇḍala 3 in the period suggested by the Ṛgveda, but he merely ‘recorded’ for that era a kind of interview of his spiritual predecessors. The prayers he put together were not his own but of these predecessors, and thus he named them as the original composers. In doing so he updated them to the local dialect, and possibly modified some geography as well. This resolution is in violation of Talageri’s interpretation of the Ṛgveda. To him, later modifications only involved changing the language of the hymns and not their historical or geographical content. This is largely true, for Sudās’ gāthina was placed exactly where is suggested by the geography of maṇḍala 3.

But equally is it true that the ideas behind these hymns, and the philosophy they were premised on, originated with Viśvāmitra Kauśika in a much earlier era. Sudās’ gāthina brought those ideas to the Sarasvatī and put together the maṇḍala, opting not to take credit for this. We must also, at least to be comprehensive in covering the field of speculations, concede that the Ṛgveda was primarily an instrument of sound. The syllables and words used in it were appropriated for their phonology and specific fitment into desired meters. We may take repeat instances of Sudās or Viśvāmitra as historical, but it’s possible that to the composers these words were used only for their audio content, their frequency evidence of their sonority. Clearly this harks back to the original multi-faceted nature of Saṁskṛta and the Ṛgveda, both.

In other words, the conflict between Ṛgvedic and Paurāṇika evidence is resolved by understanding that while the Ṛgveda is an oral interview recorded at a fixed moment in time, the Purāṇas are a hefty biography compiled at the end of life.

To take the ‘interview’ as evidence of anything other than that fixed moment is to misunderstand the nature of both the Vedas and the Purāṇas. The Ṛgveda we possess is in fact multiple recordings over previous layers. Raw sūktas existed before they were put into maṇḍalas, and even the maṇḍalas were edited and recompiled before the final tape recording. At each layer our chronology of composition and assembly should be modified accordingly, but all we possess is the singular final recording. This is already true of maṇḍalas 9 and 10, which are accepted to be part of the tape-recording yet constitute a later compilation of prayers in a somewhat different dialect.

The second example of Ṛgvedic confusion is through Bharadvāja Bārhaspatya. Paurāṇika tradition places him as contemporary to Bharata, and his descendant Vidathin is adopted by the latter. Bharadvāja Bārhaspatya is primary composer of maṇḍala 6 of the Ṛgveda. Like Sudās is the chief Bhārata of book 3, Divodāsa is the chief Bhārata of book 6. But Divodāsa is many generations after Bharata, so how would Bharadvāja Bārhaspatya have composed prayers for Divodāsa, if he lived in an earlier era? In the Purāṇas we find three Divodāsas. The first two are born in the Kāśī line, and have established synchronisms with kings in Aikṣvāku and Haihaya lines. The third Divodāsa is born in the Bhārata line at least 22 generations after Divodāsa II of Kāśī, and in the Ṛgveda the former is who we associate Bharadvāja with. To unravel Bharadvāja then, we must unravel Divodāsa first.

For our purpose we may ignore Divodāsa I of Kāśī in 3500 BC. We know that Āṅgirasas existed before him, through their association with Māndhātṛ, but they don’t particularly confuse matters for us at this stage. Problems begin with Divodāsa II, who along with his son Pratardana, and the Aikṣvāku king Sagara, mounted a notable resistance against the Haihaya conglomerate in ~3000 BC. Contemporary to this Divodāsa are the Bhārgava ṛṣis Gṛtsamada Śaunaka and Kurma Gārtsamada, primary composers of maṇḍala 2 in the Ṛgveda. RV 2–19–6 by Gṛtsamada praises Indra for helping Divodāsa destroy the cities of Śaṁbara. Like Indra, Śaṁbara is both mythic and historical. We’ve seen how the Aikṣvākus Raghu fought an Indra and Daśaratha fought a Śaṁbara, both during the same period when Ṛgvedic Bhāratas were weaving these names into hymns. Paurāṇika tradition at many places tells us that Divodāsa II and Sagara, apart from fighting Haihayas, had to reclaim their kingdoms from Rākṣasa tribes. The Śaṁbara that Divodāsa II fought was such a Rākṣasa leader, and Gṛtsamada Śaunaka weaved it into his poems.

At many places in the Ṛgveda, Divodāsa is presented under the epithet of Atithigva. Scholars agree that there were multiple Atithigvas, and that the name was used as an epithet for several Bhārata rulers. Note that a cakravartin to appear later in the Bhārata line, but much before Divodāsa and Sudās Bhārata, is named Suhotra Ātithina. One possibility is that Atithigva is the same as Ātithina, though we do not find in Suhotra’s ancestry an Atithina or Atithigva. The only possible fit is Vidathin, who’s also represented as Vitatha, but by now the cognacy is highly tenuous. At places such as RV 1–112–14, both Griffith and Jamison-Brereton translate the first line as “Atithigva and Divodāsa,” indicating they were different people. This acquires a new dimension when we understand the meaning of atithigva- whom the guests should go to.

An atithigva is the patron of a Ṛgvedic ritual, the donor, and the one on whose behalf the Devas are invoked. Indeed in manḍala 7 it’s Sudās who’s the atithigva, though he is not referred to as such. This is why there can be multiple atithigvas, and in the case of Divodāsa Bhārata it’s not always him. Contemporary to him there are Aikṣvāku kings who too are atithigvas, especially when they campaign alongside Divodāsa. And preceding him by many generations was Divodāsa II of Kāśī, who was an atithigva though the word had not come into usage yet.

Thus we should not take it as a neat equation that Atithigva = Divodāsa Bhārata. Yes, it’s primarily him, since he’s the patron of Bharadvāja’s prayers. But it’s also anyone else Bharadvāja refers to as a past invoker or beneficiary of the Devas’ beatitude.

Atithigva finds frequent mention in maṇḍala 1, in prayers composed by Savya Āṅgirasa, Kutsa Āṅgirasa, Kakṣivān Dairghatāmasa and Parucchepa Daivodāsi. Other primary composers of maṇḍala 1 are Dīrghatamāsa Aucathya and Gotama Rāhūgaṇa, making it clear that this clan of Āṅgirasas are of an early generation. Divodāsa is mentioned again in maṇḍala 4 (as Atithigva) by Vāmadeva Gautama, and in maṇḍala 7 by Maitrāvaruṇi Vasiṣṭha. To Talageri, only Bharadvāja Bārhaspatya in maṇḍala 6 speaks of Divodāsa in a contemporary sense, and all other books look to Divodāsa in the past tense. This would mean that the Āṅgirasas of maṇḍala 1 are posterior to Bharadvāja, a situation at complete odds with Puraṇic genealogy. There is then the matter of Parucchepa Daivodāsi, composer of hymns 127 to 139 of maṇḍala 1. This is likely a descendant of Divodāsa Bhārata, though curiously Parucchepa and his disciples joined the Bhārgavas, and not the Āṅgirasas. We also find the name Pratardana Daivodāsi as composer of RV 9–96, and Paurāṇika genealogy tells us that Pratardana was the son of Divodāsa II of Kāśī.

Talageri points out that the epithet Atithigva frequently occurs with another- Kutsa. To him, Atithigva is used in the Ṛgveda to refer to Bhārata kings while Kutsa refers to Aikṣvāku kings. This situation is relevant for both the Divodāsas of our concern. Divodāsa II of Kāśī campaigned alongside Aikṣvākus such as Bāhuka and Sagara, while Divodāsa Bhārata likely fought with Daśaratha at his side. In either scenario history would remember an Atithigva-Kutsa alliance take on Dasyus/Asuras/Dānavas.

The correct situation with Divodāsa(s) and Bharadvāja then is this:

• The first Divodāsa that featured in hymns was Divodāsa II of Kāśī, who is also referred to as Atithigva though he is not a Bhārata. He dates the earliest, but his hymns were ‘(re)composed and assembled’ into Ṛgveda at a later period. He featured in the hymns of Gṛstamada Bhārgava and Kakṣivān Dairghatamasa.
• The second Divodāsa to feature was the Bhārata of maṇḍala 6, patron to the historical Bharadvāja Bārhaspatya- who was neither the adopted son of Bharata nor a direct son of Bṛhaspati. He simply couldn’t have been contemporary both to Bharata and to Divodāsa. This Bharadvāja is the definitive Ṛgvedic founder- the first to put hymns into the meter, verse and dialect characteristic of the Ṛgveda. He sets the ball rolling for consequent ṛṣis on the Sarasvatī to compose and assemble prayers from their scholarly predecessors- Vāsisṭḥas, Vaiśvāmitras, Āṅgirasas and others. This is why maṇḍala 6 is chronologically the first maṇḍala, though other books feature ṛṣis that predate Bharadvāja Bārhaspatya.
• The Vidathin that Bharata adopted was a descendant of Bṛhaspati, leading to confusion later that he was Bharadvāja, as the latter too was a Bārhaspatya. Bharata adopted an early Āṅgirasa, most likely a nephew of Dīrghatamāsa, but not a Bhāradvāja. The patronymic of Bārhaspatya for Divodāsa Bhārata’s Bharadvāja could be a kind of error similar to Gāthina for Viśvāmitra Kauśika.
• To Vāmadeva Gautama of maṇḍala 4, who is an Āṅgirasa more than a century after Bharadvāja, the names Divodāsa, Atithigva and Śaṁbara are already part mythic, leaving no room for historical parsing. Further, his is the era of Ṛgvedic perfection, when the science of sound and meter founded by Bharadvāja finds elegant expression in Vāmadeva’s hymns. In his prayers these names could exist more for phonology, sonority and deeper meaning.
• Other references to Divodāsa are either historical, like by Maitrāvaruṇi Vasiṣṭha in maṇḍala 7, or meta-mythological, like by Amahīyu Āṅgirasa in maṇḍala 9.

To summarise, on the count of composition and assembly of the Ṛgveda, there are two ways to think of chronology. The first follows the broad chronology of Talageri, commencing with maṇḍala 6. This chronology is true for an assembly of the Ṛgveda as well as the final ‘expression’ in Ṛgvedic dialect and ritual. But the composition draws from pre-Vedic ideas, knowledge and ṛṣi lineages, such that past prayers were ‘composed’ afresh in the Ṛgvedic era.

The chronology of maṇḍalas is thus better expressed such:

1. Maṇḍala 6: Composed by Bharadvāja Bārhaspatya and his immediate descendants. This features their original compositions, but they draw from a long line of Āṅgirasa ṛṣis. It sets the example for Ṛgvedic composition.
2. Maṇḍala 3: Composed by Sudās’ Vaiśvāmitra gathina, who put together the hymns of his spiritual ancestors into a contemporary dialect to find prominence and royal patronage. The dialect he followed was set by Bharadvāja in maṇḍala 6, the only Ṛgvedic book likely to exist before Sudās. The Vaiśvāmitra gāthina followed the example set by royal priests of Sudās’ ancestors to gain his favour. He credited as composers the people whose ideas he expressed in Ṛgvedic form, and those original composers were Viśvāmitra Kauśika and his immediate descendants from a previous era.
3. Maṇḍala 7: Composed by Maitrāvaruṇi Vasiṣṭha, who was contemporary to Sudās Bhārata. It features original compositions that draw from a profoundly ancient lineage of Vāsiṣṭha wisdom. The precedence for Ṛgvedic composition was set by Bharadvāja, and later ṛṣis who sought the favour of Bhārata rulers had to express in the same style.
4. Maṇḍala 4: Composed by Vāmadeva Gautama, a few generations after Sudās. These are original compositions that draw on Āṅgirasa wisdom, representing a perfection of Ṛgvedic expression in the school that originated it all.
5. Maṇḍala 2: Assembled into contemporary dialect by a Bhārgava, possibly Somāhuti or Kurma, using prayers of Gṛtsamada Bhārgava from a much older time. Neither Somāhuti nor Kurma feature as Gṛtsamada’s direct descendants, so Kurma’s patronymic of Gārtsamada is distant not immediate. This marks the entry of Bhārgavas into the school of Ṛgvedic expression.
6. Maṇḍalas 5, 1 and 8 represent an extended period of co-flowering. Maṇḍala 5 largely features original compositions by Ātreyas building on the wisdom of their lineage. They were in close interaction with the Kāṇvas and Āṅgirasas of maṇḍala 8, which too features original compositions. Maṇḍala 1 is an eclectic mix of prayers by Āṅgirasas, Agastyas, Kāṇvas and Vaiśvāmitras. It shows many signs of being assembled at a later stage, and appears part of a deliberate design of the Ṛgveda. They likely comprise a period of 3–4 centuries between Rāma Dāśarathi and Hiraṇyanābha in his line, and under the latter the three Vedas Ṛg, Yajur and Sāman had acquired some kind of form.
7. Maṇḍala 9: A book of prayers to soma that makes evident why we should think of Ṛgvedic chronology in two ways. The Soma Pavamāna features composers such as Kavi Bhārgava, Kaśyapa Asita, Jamadagni Bhārgava, Madhucanda Vaiśvāmitra, Cakṣu Mānva and Kaśyapa Marīcā among others. These are names from wildly different eras, and their hymns represent the crystallisation of past knowledge into a contemporary idiom and culture. This was likely compiled in the two centuries between Hiraṇyanābha Aikṣvāku and Brahmadatta Pāñcāla.
8. Maṇḍala 10: A later addition to the tape-recording, for its language is of a different nature than that of the previous nine maṇḍalas. This could be Veda Vyāsa’s contribution, a book where he crystallises knowledge from personalities as varied as Trita Āptya, Nābhānedīṣta Mānva, Sudās Paijvana, Pṛthu Vainya and Yama Vaivasvat to name a few.

The above is in line with tradition given my nested model of Mahāyugas. Tradition records that Vedic ritual, values and mantras were instituted in the minor Tretā Yuga, the period between Sagara Aikṣvāku and Rāma Dāśarathi. In the beginning of this era are placed the first Āṅgirasas- Bṛhaspati, Dīrghatamāsa and sons- and towards the end we find the Daivodāsa Bhārgavas descended from Divodāsa Bhārata. The Dvāpara Yuga that begins with Rāma is considered the main period for compilation of Ṛgvedic maṇḍalas, and in this era we find names such as Sudās, Somaka, Hiraṇyanabha and Brahmadatta (patrons of Ṛgvedic compilation) and the authoritative Veda Vyāsa. This allows us to understand why the Vedic period began many centuries before Bharata, even during the eras of Viśvāmitra Kauśika and Jamadagni Bhārgava, while true Ṛgvedic period must await the rise of Bharata’s distant descendants.

In conclusion, the broad Ṛgvedic period ranges from around 3300 BC to 2000 BC, where composition of raw sūktas commences at the beginning and final assembly of the maṇḍalas and Veda itself freezes after 2000 BC with Veda Vyāsa. Names from the earliest stages are Māndhātṛ, Viśvāmitra and Jamadagni. Names during the middle period are Divodāsa, Bharadvāja, Sudās, Vasiṣṭha and a second Viśvāmitra. Following these are Vāmadeva Gautama, Sahadeva, Somaka and a range of ṛṣi families- Atri, Kāṇva, Bhārgava. The closing stages feature Devāpi, Śāntanu, Parāśara and finally, Veda Vyāsa. Right from the start, and at each stage, the knowledge being embedded into the Veda could date from much earlier periods- for which there is no way to discern an origin-point.