Pramāṇa · Search · About · API

Ancient Indian Historical Tradition c20: Chapter XV. The Four Ages, Chronology and Date of the Bhārata Battle

Ancient Indian Historical Tradition c20

Language
English

Contents

Chapter XV. The Four Ages, Chronology and Date of the Bhārata Battle

c20.2

TIME that is treated as historical in tradition is divided into four ages (*yuga*), the Kṛta (or Dharma or Satya), Tretā, Dvāpara and Kali (or Tiṣya), and this reckoning appears to have an historical basis, though later speculations elaborated it into an amazing yet precise scheme of cosmogony. That scheme does not render this reckoning of four ages unworthy of attention, because the genealogies refer to them sometimes, and it appears that they did correspond to certain periods.

c20.3

It is noteworthy that this theory of the four ages did not apply to the whole world. It is declared repeatedly that these ages prevailed in India (*Bhārata varṣa*),¹ and the descriptions of the other continents (*varṣa*) say nothing about the ages occurring there,² and portray conditions incompatible therewith. The four ages therefore concerned India only, and it is declared that they prevailed only in India.³ The position of these ages in the seventy-one four-age periods which made up a *manvantara*⁴ in the cosmological scheme was therefore a later elaboration. The early idea was that the four ages were a peculiarity of India alone, hence obviously the explanation of them must be sought in the conditions of ancient India.

c20.4

¹ MBh vi, 10, 387. Vā 24, 1; 45, 137; 57, 22. Bd ii, 16, 68-9; 29, 23. Mat 142, 17. Br 27, 64. Pad i, 7, 3.

c20.5

² The ages do not obtain in Plakṣadvīpa; Vā 49, 22; Bd ii, 19, 24.

c20.6

³ So Br 19, 20. Viṣ ii, 3, 19. Lg i, 52, 32.

c20.7

⁴ Vā 45, 69 and Mat 114, 1 suggest that even the theory of the fourteen Manus applied only to India. Vā 45, 67 and Mat 113, 78 have *Bhārata yuga*.

c20.8

It is a commonplace of history that great wars, conquests or political changes put an end to one age and usher in a new age, or mark the transition from one to the other; and so the Mohammedans and the British introduced new ages into India. It is natural therefore to surmise that similar changes occurred and were so regarded in ancient India, and indications of this are found in tradition.¹ The end of the Dvāpara age was admittedly marked by the Bhārata battle, for it is declared that the battle occurred in the interval (*sandhyā*) between the Dvāpara and Kali ages;² but this was afterwards modified,³ and the beginning of the Kali age was fixed at the passing away of the great heroes of that battle, Kṛṣṇa and the Pāṇḍavas, in order apparently to obviate the repugnant idea that the deified Kṛṣṇa lived into the Kali age, according to the express statement that that age began immediately he died.⁴ The broad fact however is clear, that the Dvāpara age closed with that battle, and that the Kali age began with the changes in the political condition of N. India that ensued.

c20.9

¹ This is developed otherwise, MBh v, 131, 4473-8 : xii, 69, 2693-5.

c20.10

² MBh i, 2, 282. But sometimes it is said the Kali age had already begun before the battle; e. g. MBh vi, 66, 3012 : ix, 61, 3364 ; and in the curious tale of the sleeping Mucukunda, Hv 115, 6483 ; Viṣ v, 24, 5; Br 197, 5.

c20.11

³ e. g. Kūr i, 27, 8.

c20.12

⁴ Mat 273, 49-50. Vā 99, 428-9. Bḍ iii, 74, 241. Viṣ iv, 24, 35, 36, 40 : v, 38, 8. Bhāg xii, 2, 29, 30, 33. Br 212, 8. See p. 53. Mr. Jayaswal treats this as a chronological basis, *Journal*, B. and O. Research Socy. iii, pp. 254 f.

c20.13

Tradition speaks also of an earlier time of great destruction and misery, when the kṣatriyas were well-nigh exterminated and North India was plunged into grievous calamities,⁵ and brahmanic fable attributes that to Rāma Jāmadagnya, though kṣatriya tradition shows it really occurred in consequence of the devastating raids of the Haihayas, from whom Sagara delivered the land and restored peace (chapter XXIV). That time may naturally have marked the transition from one age to another. There is no later similar period of calamity that suggests itself as a change of age, but tradition treats Rāma's destruction of Rāvaṇa and the Rākṣasas of the Dekhan and Ceylon as an epoch of signal vengeance upon evil foes.⁶ The table of genealogies in chapter XII comprises all the kings from the beginning of the Kṛta to the end of the Dvāpara age, and shows that Sagara's destruction of the Haihayas and Rāma's reign divide the whole period into three parts of not very unequal length, which might well constitute three ages. This arrangement produces a scheme of four ages which is sensible, though exactitude cannot of course be expected.

c20.14

⁵ MBh xii, 49, 1775-89 : also iii, 117, 10201-5.

c20.15

⁶ Hv 42, 2327. Br 213, 126. Gar i, 215, 8-9.

c20.16

This division accords with what tradition says about the transition from one age to another. Kṛṣṇa lived at the time of the Bhārata battle and the close of the Dvāpara age.¹ Rāma Dāśarathi lived in the interval between the Tretā and Dvāpara ages.² To Rāma Jāmadagnya is assigned the same position, and the references say he lived in the Tretā age,³ and smote the kṣatriyas in the interval between the Tretā and Dvāpara ages.⁴ But this was Rāma Dāśarathi's position, and that particularization is clearly wrong, for Rāma Jāmadagnya was avowedly prior as shown by the synchronisms in chapter XIII, and the allegation that he destroyed all kṣatriyas off the earth twenty-one times (really the long-continued Haihaya devastations) is wholly incompatible with the story of Rāma Dāśarathi. It is obvious that Rāma Jāmadagnya belonged to the interval between the Kṛta and Tretā ages, when in fact the Haihayas had their dominion, and the references should be to the Kṛta age and that interval. The Kṛta age then ended with the destruction of the Haihayas; the Tretā began approximately with Sagara and ended with Rāma Dāśarathi's destruction of the Rākṣasas; and the Dvāpara began with his reinstatement at Ayodhyā and ended with the Bhārata battle: so that, taking the numbers in the table of genealogies, the division is approximately thus, the Kṛta Nos. 1–40, the Tretā Nos. 41–65, and the Dvāpara Nos. 66–95.

c20.17

¹ MBh xii, 341, 12953–4 : cf. vi, 66, 3012.

c20.18

² MBh xii, 341, 12949.

c20.19

³ MBh xii, 341, 12948.

c20.20

⁴ MBh i, 2, 272; yet 64, 2480 absurdly makes the Kṛta (read Tretā) age begin after Rāma's devastation (2459). Hv 106, 5869. Gar i, 215, 7 is nearly right.

c20.21

These considerations show how the belief arose that Viṣṇu became incarnate when conditions on earth had become evil, in order to destroy wickedness and re-establish righteousness,⁵ for his three chief alleged historical incarnations were, the earliest Rāma Jāmadagnya, the second Rāma Dāśarathi, and the last Kṛṣṇa. Such statements are brahmanical, and historical consistency cannot be expected in all.⁶ The misstatement of Rāma Jāmadagnya's position arose probably from the notion that the Kṛta age was one of unblemished righteousness.

c20.22

⁵ Vā 96, 232; 97, 64–6; 98, 69, 98. Bḍ iii, 73, 70, 97. Mat 47, 235, 247. Br 180, 26–7; 181, 2–3.

c20.23

⁶ So it seems to be said the Pāṇḍavas lived in the interval between the Tretā and Dvāpara ages, MBh iii, 121, 10310; 125, 10409.

c20.24

There are allusions to the ages sometimes in the genealogies, and these (when without the elaboration of the ages which will be noticed) appear to be sometimes right, and tend to show that there was in kṣatriya tradition some memory of the chronological position of certain important events. Thus king Bāhu of Ayodhyā lived in the Dharma (Kṛta) age,¹ which agrees with the position, No. 39, determined for him. Karandhama of the Vaiśāla dynasty reigned at the beginning of the Tretā age,² and his twelfth successor Tṛṇabindu reigned at the third mouth of that age,³ which appears to mean the beginning of the third quarter of it: and the positions, Nos. 38 and 52, determined for them practically agree with the approximate limits assigned to that age above. But most such allusions occur in stories and discourses, often brahmanic, and are sometimes right but more often wrong.⁴ There is an inclination to assign events to the Tretā age,⁵ and the expression Tretā-yuga⁶ means at times little or nothing more than ‘once upon a time’. Such statements are generally worthless for chronological purposes.

c20.25

¹ Bḍ iii, 63, 121. Vā 88, 123. Hv 13, 761, Br 8, 30. Śiv vii, 61, 23.

c20.26

² Vā 86, 7. Or his son, MBh xiv, 4, 80.

c20.27

³ Bḍ iii, 8, 36–7; 61, 10–11. Vā 70, 31; 86, 15.

c20.28

⁴ e. g. Vā 30, 76: Bḍ ii, 13, 83: Mārk 7, 1: MBh xiii, 14, 701–2; 150, 7128.

c20.29

⁵ e. g. Vā 8, 201; 9, 46; 30, 76; 57, 39, 43; 91, 48.

c20.30

⁶ e. g. Br 34, 48: Vā 67, 43.

c20.31

It is unnecessary here to pursue this matter into the later fully developed theory of the yugas and manvantaras, wherein 71 four-age periods (catur-yuga) made up a manvantara. It was a fanciful brahmanical elaboration; and one feature in it is that the present time is the Kali age in the 28th four-age period of the Vaivasvata manvantara, so the events of traditional history were sometimes distributed among those 28 periods.⁷ Thus a pretentious passage declares⁸—Datta Ātreya as Viṣṇu’s fourth incarnation and Mārkaṇḍeya lived in the 10th Tretā age (i. e. in the Tretā age of the 10th four-age period); Māndhātṛ as his fifth incarnation⁹ and Utathya lived in the 15th Tretā; Rāma Jāmadagnya as his sixth and Viśvāmitra lived in the 19th Tretā; Daśaratha’s son Rāma as his seventh and Vasiṣṭha lived in the 24th age¹; Vyāsa as his eighth with Jātūkarṇya, and Kṛṣṇa as his ninth with Brahma-Gārgya lived in the 28th Dvāpara.² Such assignments sometimes observe some chronological consistency, often they are erratic, and in any case, being brahmanical notions lacking the historical sense, they are unreliable.

c20.32

⁷ A short explanation will be found in Hastings’s Dict. of Religion and Ethics, s.v. ‘Puranas’.

c20.33

⁸ Vā 98, 88–97. Bḍ iii, 73, 87–97. Mat 47, 242–6.

c20.34

⁹ Interpolated in Vā 88, 69. Not in Bḍ.

c20.35

¹ So also Bḍ iii, 8, 54: Vā 70, 48: Br 213, 124. Yet inconsistently, Rāvaṇa lived in the Tretā in the second period, Br 176, 15–16: and both are placed in the 27th period, Pad v, 14, 67–8.

c20.36

² In the 27th, Pad v, 23, 7–9.

Date of the Bhārata Battle

c20.s1.1

As the Bhārata battle marked the end of the Dvāpara age and is a great landmark, it is well to reach some estimate of its probable date.³

c20.s1.2

³ In the following discussion I refer for convenience to my *Dynasties of the Kali Age*, as DKA.

c20.s1.3

Candragupta began to reign in or about 322 B. C. He was preceded by the nine Nandas,⁴ Mahāpadma and his eight sons, who are said to have enjoyed the earth one hundred years. To Mahāpadma are assigned 88 years and to his sons 12 years. The best reading says, not that he reigned 88 years, but that he would be (that is, lived) 88 years; and a hundred years for the joint lives of him and his sons accord with an ordinary genealogical estimate, and are not unreasonable, as his life was long. It is improbable in the circumstances of that time that he could have gained the throne of Magadha until he was grown up, or, say, 20 years old at least.⁵ The reigns of the nine Nandas would then be reduced to 80 years, and we may reckon that they began approximately at (322 + 80) 402 B. C.

c20.s1.4

⁴ DKA, pp. 25–6, 69–70.

c20.s1.5

⁵ A variation in this estimate makes no material difference.

c20.s1.6

The next question to consider is the time between Mahāpadma’s inauguration and the Bhārata battle. For this three sets of data are alleged. First, there reigned in Magadha during that time 22 Bārhadrathas, 5 Pradyotas and 10 Śiśunāgas, and the total of all their reigns is (940 + 138 + 330) 1408 years, while the totals of the durations of the dynasties vary from (1000 + 138 + 360) 1498 to (723 + 52 + 163) 938 years according as we take all the highest or all the lowest figures.⁶ Secondly, it is said that the period from Mahāpadma’s inauguration back to Parikṣit’s birth, which occurred soon after the battle,¹ was 1050 (or 1015) years.² These figures are so discrepant that it is clear no reliable tradition has survived in them. It cannot be said that any one of them is more trustworthy than the others. No calculation can be based on all of them combined, and to make computations from one or other of them is purely random work.³ Besides they are all demonstrably wrong. From the Bhārata battle to Mahāpadma there were 30 Paurava kings (for Yudhiṣṭhira must be reckoned in) and 29 Aikṣvākus (excluding Siddhārtha, i. e. Buddha, who did not reign), beside the 37 Māgadha kings; hence on a reckoning of the kings as 30, the foregoing figures, 1408, &c., give average reigns of 47, 50, 31 and 35 years respectively, which are all impossible when tested by real historical averages as will be shown. Those figures therefore cannot be relied on. The third set of data is that Mahāpadma exterminated all kṣatriyas, and that until then there reigned contemporaneously for the same length of time 24 Aikṣvākus, 27 Pañcālas, 24 Kāśis, 28 Haihayas, 32 Kaliṅgas, 25 Aśmakas, 36 Kurus, 28 Maithilas, 23 Śūrasenas and 20 Vītihotras.⁴ Here we have safer ground, for the (names and so) number of kings in a dynasty was a much simpler matter and more easily remembered than figures of the lengths of reigns and dynasties; and this information about ten contemporary dynasties eliminates peculiarities and extravagances about single dynasties and enables us to make prudent calculations by means of averages of all ten. The investigation will proceed on these lines.

c20.s1.7

⁶ DKA, pp. 13–22, 67–9.

c20.s1.8

¹ MBh xiv, 66 to 70.

c20.s1.9

² DKA, pp. 58, 74.

c20.s1.10

³ Mr. Jayaswal in *Journal, B. and O. Research Socy.* i, pp. 67 f. : iii, pp. 246 f. : iv, pp. 26-35. The astronomical statements obviously cannot have scientific precision, and can only have been formed by estimate at the close. Very probably regnal years have been unduly swelled by reckoning for a king his *yuvarāja* period as well as his reign proper.

c20.s1.11

⁴ DKA, pp. 23-4, 69.

c20.s1.12

It would have taken Mahāpadma some time to conquer all those kingdoms, the nearer earlier and the more distant later; and it will not be far out if we strike a mean, say, of 20 years after his accession for their destruction,⁵ and so fix the year (402-20) 382 B. C. for their mean termination. That list of contemporary kings can be tested as regards its period and the number of kings. The dynastic account gives the Paurava, Aikṣvāku and Bārhadratha kings from the time of the battle, but the prophetic portion of it starts from the point of time when the Paurava king Adhisīmakṛṣṇa, the Aikṣvāku Divākara and the Bārhadratha Senājit were reigning contemporaneously, for it is clearly stated in the Paurava list that the future kings were 25, Adhisīmakṛṣṇa and his 24 successors, and in the Bārhadratha list that they were 16, Senājit and his 15 successors; the predecessors in both cases not being so reckoned. Hence in these three dynasties that point of time is the real initial point throughout and the extermination by Mahāpadma is the final point.

c20.s1.13

⁵ This estimate may be varied without material difference.

c20.s1.14

The number of kings can be tested as regards the Aikṣvākus and the Kurus who were the Pauravas. The Aikṣvāku list names 25 future kings from Divākara (omitting Siddhārtha), and the list of contemporary kings says 24 Aikṣvākus, so that the two agree practically. The Paurava-Kuru list names 25 future kings, and the contemporary list says 36; but another well-attested reading in the latter says 26,¹ and it was pointed out² that, because of the ease with which *tr* and *v* might be confused, ‘in many cases either [20 or 30] may be read as other data may indicate, irrespective of the weight of the MSS.’ This reading 26 is no doubt the true reading, because it accords better with the other numbers in the contemporary list and agrees practically with the 25 in the former list. The practical agreement in these two dynasties, the only cases we can test, indicates that the contemporary list is also reckoned from the same initial point as the three detailed dynasties.

c20.s1.15

¹ DKA, p. 24, note 16.

c20.s1.16

² Id, p. xxiii, § 39.

c20.s1.17

According to the contemporary list then there reigned between those initial and final points, 24 Aikṣvākus, 27 Pañcālas, 24 Kāśis, 28 Haihayas, 32 Kaliṅgas, 25 Aśmakas, 26 Kurus (Pauravas), 28 Maithilas, 23 Śūrasenas and 20 Vītihotras,³ that is 257 kings in ten kingdoms, or a mean of 26 kings. For these 26 then we must allow reigns of medium length, and the question is, at how many years should a medium length be reckoned? The longest average of reigns occurred among the 20 Vītihotras, the shortest among the 32 Kaliṅgas. Thus 20 long reigns = 32 short reigns = 26 medium reigns, whence we obtain the proportion—longest average : shortest average : medium average :: 26 : 16¼ : 20. I have examined 14 series of from 20 to 30 kings in various eastern and western countries; the longest average just exceeded 24 years in one case, the shortest was about 12, and the average of all was 19; but the average was higher in western countries and lower in eastern countries. Hence as a medium average for these contemporary eastern dynasties we must take something less than 19, and 18 years will be a fair and even liberal estimate. The duration of these ten contemporary kingdoms then would be 26 × 18, that is, 468 years, and their period would be from 850 to 382 B. C. when Mahāpadma exterminated them.

c20.s1.18

³ It is said the Vītihotras had passed away before the Pradyotas began, DKA, pp. 18, 68. If so, they should be omitted; yet the reckoning here would not be materially modified.

c20.s1.19

In this calculation the Magadha kings have not been included, since they are omitted from the contemporary list, and the date 850 B. C. may now be tested with reference to them. From Senājit (850) till Mahāpadma overthrew the Śiśunāgas (402) reigned 16 Bārhadrathas, 5 Pradyotas and 10 Śiśunāgas; that is, 448 years are allowed for 31 reigns—an average of 14½ years. This lower average is quite probable because of the violence that overthrew those dynasties, and it is about the average I have found in eastern dynasties. The above estimate therefore of 18 years for a medium peaceful reign appears just, and the date 850 B. C. is highly probable.

c20.s1.20

This year 850 would be the approximate mean date of the beginnings of the reigns of Adhisīmakṛṣṇa, Divākara and Senājit; and therefore the standpoint during their reigns, dividing the ‘past’ from the ‘future’ in the prophetic account,¹ would be a few years later, say, about 840 B. C.

c20.s1.21

¹ See Vā 1, 12–15; 99, 258–9, 282, 300; Mat 1, 4–5; 50, 66–7; 271, 5, 23: and p. 52.

c20.s1.22

To get the time of the Bhārata battle, we must add the kings who preceded those three kings, namely, 5 Pauravas (for Yudhiṣṭhira’s reign must be included), 4 Aikṣvākus and 6 Bārhadrathas, that is, a mean of 5, and here for so short a period the medium reign probably was longer, say 20 years.² Hence we must add (5 × 20) 100 years, and the date of the battle may be fixed approximately as (850 + 100) 950 B. C.³ This reckoning has avoided special figures or unique statements, except in the one case of the Nandas where no other course is available; and has proceeded upon general averages (1) of the number of kings belonging to 11 contemporaneous dynasties and (2) of the lengths of reigns computed from the reliable data of 14 historical dynasties in other countries; thus it has eliminated all peculiar features and is a reasonable general approximation.¹

c20.s1.23

² The 60 years assigned to Parikṣit II cannot be relied on (p. 53).

c20.s1.24

³ Mr. Jayaswal fixes the battle in 1424 B. C., and other Indian writers favour similar early dates, all working on the above chronological statements in the Puranas (which are discrepant) without checking their figures by comparison with reliable data from dynasties elsewhere. Such a comparison shows that their calculations produce results contrary to general experience: thus his date makes the medium average of 31 reigns from the battle to Mahāpadma about 33 years, an incredible length.

c20.s1.25

¹ With a possible error of being too liberal. For further calculations see end of chap. XXIV.

c20.s1.26

If we should seek to make an estimate of the ages before the battle, it would be prudent to take a smaller length for the average reign, because only one line, that of Ayodhyā, is practically complete, while there are gaps in the other dynasties, so that there is little scope for taking medium averages of all the dynasties and eliminating peculiarities. The lowest average mentioned above, 12 years, therefore would be a sounder estimate. Since however it is said that insignificant kings have been omitted from the genealogies (p. 89), it may be contended that the average should be increased to compensate for lost kings, say, 13 or 13½ years per step in the table in chapter XII; but the uncertainty and peculiarity in such details require caution, and it would be more prudent to adhere to 12 years as the average. To contend for immense antiquity for the earliest ages is discredited by the historical sense; and to push back the antiquity of those ages to vast figures is to weaken *pro tanto* the trustworthiness of tradition about them when everything depended on memory alone.²

c20.s1.27

² Indian writers are prone to do this; e. g. Abinas Chandra Das in his *Rigvedic India*.

c20.s1.28

Another consequence of such dating is to prolong also the ages before the battle; and to put back the antiquity of any event is to weaken the trustworthiness of tradition about it.

Linked texts (13)