Chronology and the Upanishads – Koenraad Elst

Dr. Koenraad ElstAccording to Hindu tradition, the grandfather of the Mahabharata heroes is the one who ordered the Vedas in their definitive and still prevailing form: Krishna Dvaipayana alias Veda-Vyasa, … In the oldest Upanishads already, the Rg-, Sama- and Yajur-Veda are mentioned as known and complete wholes. So, the Upanishads started at or after the time of the Mahabharata battle, though at a different location. … The first six or so of the Upanishads easily predate the Buddha. The line of thought laid down in the Upanishads thus had almost a thousand years to develop, between the Mahabharata war (ca. -1400) and the Buddha (-500). – Dr. Koenraad Elst

Vedic Rishi A famous Flemish emeritus professor recently reacted to my off-hand mention of a date for the major Upanishads, viz. “second millennium BCE”. He thought that this should be 900-500 BCE, a date obviously borrowed from the textbooks. It is no big deal that a Western philosopher, not specializing in the chronology of Indian history, should abide by the received wisdom in this matter; but the few specialists know it to be highly controversial.

When scholar upon scholar claims just as off-hand that the Brhadaranyaka, Chandogya, Katha etc. Upanishads date from 900-500 BC, I always wonder: how do they know this? Where did they get it? The texts themselves never give such a date, nor other premodern texts referring to them.

Gautama BuddhaA few scholars even date them all later than 500 BCE, the time of the Buddha. Animated by the “Hinduism bad, Buddhism good” doctrine, they are puzzled by the existence of undeniably profound ideas in the Upanishads, clearly related to the Buddhist teachings, so they want to explain these as “borrowed from Buddhism”. Of course, the notion of Buddhism as a separate religion was constructed only recently, by the first Western Buddhologists, whereas Hindu tradition rightly considers the Buddha as only one of the Hindu sect founders/leaders of his time, though for sociological reasons (his high birth and top connections) the most successful one. Hindu writers of idol-making manuals treat the Buddha on a par with Krishna and the others. At any rate, the linguistic anteriority of the preclassical Sanskrit of the Upanishads date them to before the Buddha and to his probable contemporary Panini, the codifier of classical Sanskrit. Though most Western and Westernized scholars share this “Hinduism bad, Buddhism good” framework, they still agree that the great Upanishads definitely predate the Buddha. That is why these works don’t refer to any specifically Buddhist concept while basic Buddhism does refer to generally Upanishadic concepts.

But for them, 900 BCE is more than enough time distance to 500 BC. The Upanishads should not be dated earlier, for then the Aryan invasion framework runs into difficulties. This is roughly as follows:

  • 1700-1500 BCE: Indo-European or “Aryan” nomads invade India;
  • 1500-1200 BCE: they compose the Rg-Veda;
  • 1300-500 BCE: they compose the other Vedas and their ancillary literature;
  • 900 BCE onwards: among these writings are the Upanishads.

In Chinese history, all important and numerous unimportant events are dated precisely from at least the 8th century BCE, and approximately so for a thousand years earlier. In Indian history, by contrast, many important events or the birth years of famous persons are only vaguely known, mostly but not even always in their proper chronological order, and without any absolute chronology. We maintain that the usual estimate for the first Upanishads misses the mark by at least five hundred years.

Devi Saraswati blesses Rishi YagnavalkyaThere are only few chronologically relevant references in the Upanishads, and that mostly to other insecurely dated characters of Hindu literature. Thus, Yajnavalkya of Brhadaranyaka fame wins a debate at the court of king Janaka of Videha, and this king is usually taken to be the father of the Ramayana’s heroine Sita, also known by her patronymic Janaki. But that doesn’t get us very far, if only because the paucity of data makes it difficult to be sure that the same Janaka is meant, as this is not an unusual name. But the larger literary framework contains better chronological clues.

The ecliptic was divided into 28 lunar houses, like in China and Arabia, rather than in the 12 Babylonian-Hellenistic signs of the Zodiac. The precession of the equinoxes, at 1° per ca. 71 years, makes data about the relation between fixed stars and equinoxes or solstices or other seasonal phenomena into a secure chronological pointer. For instance, an ancillary work of the Vedas, the Vedanga Jyotisha (“Veda-Ancillary of Stellar Science”, at that time meaning astronomy though now used to refer to astrology), conventionally dated to 500-200 BCE, dates itself twice to ca. 1350 BCE, viz. by explicitating which stars are on the winter solstice and spring equinox points. This is an explicitly post-Rg-Vedic texts, so the Rg-Veda was already complete by the time the Aryan Invasion Theory lets the Aryans invade India. It is quite amusing to read the mental and verbal acrobatics which conformistic scholars try out to neutralize this inconvenient evidence.

The Atharva-Veda lists all the lunar houses, starting with Krttika/Pleiades, presumably because it was on the spring equinox point, which it was ca. 2300 BCE. The Yajur-Veda also gives this position thrice. But could this not be a reminiscence, a classical enumeration which endured even when the asterism concerned had, after some 900 years, shifted and left the place of honour to the next asterism? Unlikely, for astrologers typically change the list to reflect the changing of asterisms on the equinoctial point; they no longer treat Krttika as number 1. The Shatapatha Brahmana, traditionally held to be from the same period (but which modern textbooks date to 900 BCE or so), later than the Rg-Veda but just before the Upanishads, confirms this synchronism by referring to the same position, telling us that Krttika “never swerves from the east”, i.e. from the equinox, the intersection of ecliptic and equatorial plane. This passage, like the Vedanga Jyotisha passages, is part of a practical instruction under which to conduct a certain ritual, so it is observational par excellence, not a traditional prayer-type text where an ancient reminiscence might have survived.

Death of BhishmaThe Kaushitaki Brahmana points to the same period, ca. 2300 BCE, by means of a different astronomical pointer, viz. the star Regulus on the summer solstice point. Another text where the position of Regulus is mentioned, is the epic Mahabharata, but there it is said that an event, the death of the hero Bhishma coinciding with the full moon near Regulus, took place after the winter solstice, i.e. centuries after 2300 BCE.

The importance of this information is that it shows how the astronomical evidence does not always support a high chronology. Indians say that the war described in the Mahabharata took place in 3139 BCE. This is based on the length of the Four Ages given in the Puranas, a type of mythohistorical literature from the first millennium CE. The doctrine of Four Ages is very ancient, attested also in Greek and Germanic mythology, but their quantification is apparently linked with the precession of the equinoxes, discovered by Hipparchos ca. 150 BCE and introduced in India only in subsequent centuries. There are no pre-Hellenistic mentions of the fourth age starting in the 32th century BCE, 37 years after the Mahabharata war, as Hindus traditionally (i.e. Puranically) believe. The astronomical evidence of the Mahabharata itself, however, points to well after 2300 BC.

This lower chronology is supported by another passage from the Puranas. As said, it mixes historical data with mythology, so we have to be very cautious, but the following sentence is sufficiently clear. The Puranas record that either 1050 or 1015 years elapsed between the birth of Mahabharata hero Arjuna’s grandson Parikshit and the coronation of Magadha-based emperor Mahapadma Nanda in either 378 or 382 BCE. This puts the Mahabharata war (predating Parikshit’s birth by less than a year) in about 1400 BCE. This is earlier than the textbooks’ date of 900 BC, but lower than the traditional date. It is further confirmed by archaeological periodization. The second half of the second millennium BCE marked the high tide of chariot warfare, cfr. the war between the Hittite and Egyptian empires or the Trojan war. Chariot warfare is central to the Mahabharata story, not some literary addition by a later editor. And there was simply no chariot warfare in 3139 BCE.

Rishi VyasaAccording to Hindu tradition, the grandfather of the Mahabharata heroes is the one who ordered the Vedas in their definitive and still prevailing form: Krishna Dvaipayana alias Veda-Vyasa, “analyser/editor of the Veda”. In the oldest Upanishads already, the Rg-, Sama- and Yajur-Veda are mentioned as known and complete wholes. So, the Upanishads started at or after the time of the Mahabharata battle, though at a different location. The Bharata clan was based between the Yamuna and the Saraswati/Ghaggar, in present-day Haryana, while Yajnavalkya, the philosopher featured in the first Upanishad, is shown winning a debate at the court of king Janaka in Mithila, in the present-day Bihar, far to the east. At any rate, the first six or so of the Upanishads easily predate the Buddha. The line of thought laid down in the Upanishads thus had almost a thousand years to develop, between the Mahabharata war (ca. -1400) and the Buddha (-500).

Conclusion: the astronomical evidence (not treated in its completeness here) is internally consistent, faithfully following the relative chronology of the different Vedic writings, e.g. it does not date the Upanishads earlier than the Rg-Veda. It is higher than the conventional AIT chronology, irreconcilable with it. But it is also lower than some of the wilder chronologies popular in India.

» Koenraad Elst distinguished himself early on as eager to learn and to dissent. He studied at the KU Leuven, obtaining MA degrees in Sinology, Indology and Philosophy. After a research stay at Benares Hindu University he did original fieldwork for a doctorate on Hindu nationalism, which he obtained magna cum laude in 1998. As an independent researcher he earned laurels and ostracism with his findings on hot items like Islam, multiculturalism and the secular state, the roots of Indo-European, the Ayodhya temple/mosque dispute and Mahatma Gandhi’s legacy. He blogs at

3 Responses

  1. A fan of Nilesh Oak’s date for the Mahabharata war at the mid-6th millennium BC asks me what I think of his crucial argument for this very high chronology, viz. the observation that Arundhati/Alcor rose before its twin star, Vasishtha/Mizar in the Saptarishi/Great Bear. Here is a first attempt.

    Vyasa tells Dhrtarashtra that he has seen a number of evil omens in the sky, presaging the fratricidal war between the Kauravas (Dhrtarashtra’s sons) and the Pandavas (his brother Pandu’s sons) with a terrible outcome. Among these evil omens is that he saw Arundhati rise before her twin star Vasishtha. Between the 12th and 5th millennia BCE, this was normal. Ever since, it was abnormal, as Vasistha would rise ahead of Arundhati. Therefrom, NN Oak deduces that the Mahabharata war must date back to the time when Arundhati was ahead, for instance in the mid-6th millennium.

    For two reasons, this conclusion is incorrect. Firstly, an evil omen is normally either a case of muddled visibility (e.g. in Babylonian astrology Algol, in Arabic “the Demon”, counts as a negative influence because, as modern astronomy has found out, it is a double star of which the darker member periodically obscures the brighter member; and this temporary lack of visibility is deemed a bad omen; or the constellation of Cancer is deemed inauspicious because it has no bright stars and is thus a dark pit in the starry sky) or because it is an abnormality, e.g., in all astrologies the world over, an unforeseen eclipse, or the unexpected appearance of a comet. Now, between the 12th and 5th millennia BCE it would be normal to see Arundhati rise first, and thus not fit to serve as an evil omen.

    Not having a Sanskrit Mahabharata handy, I am quoting from memory, so you may fill in relevant details. I do not know how Vyasa’s observation came about, but it cannot possibly be a sighting of Arundhati rising first before the 5th millennium. But it can very well be an unusual sighting of Arundhati after the 5th millennium. So, let’s drop this eccentric theory and discuss possible dates for the Mahabharata war within the usual range of dates.

    Secondly, note already that you put the tradition and the numerous Hindu believers in the tradition in the wrong. They, with their 3139 BCE, are said to be more than 24 centuries off. So, you may fight it out with them, not to speak of those convinced of a later date, like me. Moreover, your discovery fails to take into account the general picture, with its chronology determined by many other factors. Thus, the Kaushitaki Brahmana and the Shatapatha Brahmana are astronomically dated to ca. 2300 BCE, and are literarily deemed contemporaneous with the Yajurveda, thus earlier than the editor of the Vedas, Vyasa, the grandfather of the Mahabharata protagonists. The Mahabharata itself posits that the full moon in conjunction with Regulus/Magha took place after the solstice, so after 2300 BCE when this star passed the solstice. For a different information given by the epic itself: Balarama misses the battle because he goes on pilgrimage reaching the place where it stops flowing, thus after the drying of the Saraswati ca. 1900 BC.

    And if you are skeptical of literary and astronomical data, consider the harder material data. The central event of the MBh is a chariot battle. Now, chariot battles were popular in a very specific window of history, viz. the bronze age when metallurgy was already sophisticated enough for strong and light chariot wheels; and before the rise of cavalry made chariots redundant. The first chariot found by archaeology dates to 2200 BC (Sintashta, Asian Russia). Its high tide in warfare was in the later 2nd and the early 1st millennium BC: many wars of that period were fought with chariots, e.g. the Trojan war of ca. 1200 BC. That is why the scriptural reference dating Pandava Arjuna’s grandson Parikshit 1050 years before the coronation of Mahapadma Nanda (ca. 380 BC), i.e. in the 15th century BC, may well be right. The classical date, 32nd century BCE, is extremely unlikely to have seen chariot warfare, and for Oak’s date of the mid-6th millennium BCE, this is simply impossible.

    So, I don’t know yet how to make astronomical sense of this alleged observation of Arundhati rising before Vasishtha. But a number of different reasons exclude the explanation that it points to a date in the 12th-5th millennium bracket.


  2. What does Mr. Elst have to say about the Arundhati observation given in Mahabharata?

    How does that influence his “internally consistent” astronomical data?


  3. the last word on the ancient indian chronology is yet to be said. we are still groping in the dark and it may take a few more decades before we are clear about many things.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: