Pramāṇa · Search · About · API

Ancient Indian Historical Tradition c11: Chapter VI. The Puranas and Their Genealogical Texts

Ancient Indian Historical Tradition c11

Language
English

Chapter VI. The Puranas and Their Genealogical Texts

c11.2

THE genealogies of the ancient dynasties are the chief data of an historical kind, and the investigation of traditional history must begin with them. They are given more or less fully by all the Puranas, except the late Varāha, Vāmana, Skanda, Nāradīya and Brahmavaivarta, and the Bhaviṣya which does not deal with the ancient past. Some are found in the Mahābhārata and Rāmāyaṇa. All the accounts are in verse in the śloka metre, except some parts in the Mahābhārata and most of the Viṣṇu. The most important Puranas as regards genealogies are the Vāyu and Brahmāṇḍa, the Brahma and Harivaṃśa, the Matsya (with book V of the Padma) and the Viṣṇu.

c11.3

The Vāyu and Brahmāṇḍa have the best text of the genealogies. Their accounts agree closely, so that they are really only two versions of the same text. They have a great part of their contents in common, generally almost verbatim, and it appears they were originally one Purana. This is indicated also by the lists given in the Puranas.¹ Nearly all mention the Brahmāṇḍa, putting it last, and omit the Vāyu, though it was one of the best known. The Kūrma² also mentions it last and calls it the Vāyavīya Brahmāṇḍa. The Vāyu and Brahmāṇḍa are named separately only in the Vāyu itself and the Garuḍa.³ These lists are of course late insertions, which could not have been completed till after the latest of these Puranas was composed; yet they show that the two were not regarded as distinct and the differentiation of the one original into two versions with separate names was a later process. Both these Puranas say they were declared by the god Vāyu;¹ thus both were Vāyu-prokta and either might be so described; but the use of the name Brahmāṇḍa in the above lists to the general exclusion of the name Vāyu suggests that Brahmāṇḍa was the older and better-known name. Their version may be conveniently called the ‘Vāyu version’.

c11.4

¹ Mat 53, 11–58. Viṣ iii, 6, 22–4. Lg i, 39, 61–3. Var 112, 69–72. Pad i, 62, 2–7 : iv, 111, 90–4 : vi, 219, 25–7 ; 263, 77–84. Śiv v, 1, 38–40. Mārk, p. 659 ; my translation 137, 8–11.

c11.5

² Kūr i, 1, 13–15.

c11.6

³ Vā 104, 2–11 (Anila). Gar i, 215, 15–16 (Vāyavīya).

c11.7

¹ Vā 1, 47, 196; 2, 44. Bḍ i, 1, 36.

c11.8

The editions cited are the Ānandāśrama for the Vāyu, and the Veṅkaṭeśvara for the Brahmāṇḍa. Where they differ, the former is generally preferable, because the latter is not a critical edition, and also appears to have been silently emended by the editor, as, for instance, where it reads iti śrutaḥ and viśrutaḥ instead of iti śrutiḥ (p. 19), and probably where it avoids a difficulty by substituting yogeśvarasya for Pauroravasya (i.e. Paurūravasya).² But the Vāyu is not invariably better, because sometimes corruptions have passed undetected in it, as where the Viṣṇuvṛddhas have been misplaced,³ and where it reads Ayāsya’s name wrongly.⁴ The Brahmāṇḍa has unfortunately one very serious lacuna in its account after iii, 74, 103, where the latter half of the Ānava genealogy, the whole of that the Pauravas, and a portion of the Kali age dynasties have been lost, namely, all the matter contained in Vāyu 99, 102–290. These two appear to be the oldest of the Puranas that we possess now, and are on the whole the most valuable in all matters of traditional history.

c11.9

² Bḍ iii, 66, 74. Vā 91, 102. Cf. Hv 27, 1468; 32, 1773 and Br 10, 63. See Kānyakubja dynasty in chap. IX.

c11.10

³ Vā 88, 79. See chap. XXIII.

c11.11

⁴ Vā 59, 101 with Bḍ ii, 32, 110.

c11.12

The Brahma and Harivaṃśa agree closely in their genealogical account and have practically the same text, subject to small variations. The Brahma is cited from the Ānandāśrama edition. The Harivaṃśa text (Calcutta edition) is better than the Brahma, for the latter has suffered through losses; thus it is manifestly incomplete in the North Pañcāla genealogy, and most copies of it omit the Cedi-Magadha dynasty descended from Kuru (chapter IX). Their version is very similar to the ‘Vāyu version’, and has the same basis, and appears to be a revision of that version. Sometimes it has omissions, sometimes additions which seem to contain genuine tradition, sometimes it varies, and sometimes it makes mistakes. It may be called the ‘Harivaṁśa version’. It is manifestly later than the ‘Vāyu version’, and this is best illustrated by the fact that it gives two incompatible origins for the Kānyakubja and Kāśī dynasties, one in each case being necessarily wrong, devised at a later time when erroneous views had obtained currency.

c11.13

The genealogical record in the Matsya has peculiarities. It may be considered in three parts, (1) its account of the Aikṣvākus, Śāryātas and other sons of Manu, (2) its account of the early Ailas down to Yayāti, and (3) its genealogies of the five Aila races, Yādavas, Pauravas, &c. This third part resembles the ‘Vāyu version’, and appears to be based on the same original text, and to be not a revision but a distinct version which early became separate. Its variations are additions, omissions, condensations and sometimes corruptions; and on the whole its agreement with the ‘Vāyu version’ occurs more in the strictly genealogical statements and less in the incidental or collateral matter. The two other parts (1) and (2) differ from the ‘Vāyu version’, the verses being quite different. The pedigrees are in the main the same, though there are wide divergencies or corruptions in names in the second part and marked disagreements at several stages in the first part. In the first part the account is very concise, without any of the tales and allusions that diversify the ‘Vāyu version’. The second part is briefer than the ‘Vāyu version’, the legends are narrated differently, and some interpolations occur. On the whole the Matsya record is a valuable and in many points independent authority.

c11.14

The genealogical account in the Padma¹ is in book V, and is practically the same as the Matsya version so far as it goes, the differences being generally small and verbal. It is therefore valuable as a means of checking the Matsya text.

c11.15

¹ Ānandāśrama edition.

c11.16

The Viṣṇu account is mainly in prose, with old verses introduced occasionally. It is generally in agreement with the ‘Vāyu and Harivaṁśa versions’ in the structure of its genealogies, sometimes agreeing rather with the former as in the Aikṣvāku dynasty, and sometimes rather with the latter as in the cases of the Yādavas and Pauravas; but it also has omissions, variations, additions, and ‘embellishments’ of its own. It leaves out some of the incidents in those versions, condenses others (as for instance, the famous story of Satyavrata Triśaṅku of Ayodhyā ¹), and adds others again, which are brahmanical fables (as the story of the rishi Saubhari and king Māndhātṛ's daughters) ² or brahmanized legends (as the story of king Kalmāṣapāda),³ or tales that seem to have some basis of genuine tradition but have been half mythologized (as the story of Purukutsa and Narmadā).⁴ The verses it intersperses are manifestly quotations from older metrical versions, and agree sometimes with the Vāyu and Harivaṃśa versions. It is a late Purana, composed as a single whole upon a consistent plan, and not a collection of materials of various times, as we find in the Vāyu, Brahma and Matsya. From its account of Buddhism and Jainism (p. 68) it appears to have been composed after brahmanism had recovered its supremacy, so that it cannot be earlier than about the fifth century, A.D.⁵ and it is brahmanical.

c11.17

¹ iv, 3, 13-14.

c11.18

² iv, 2, 19 to 3, 3. P. 73.

c11.19

³ iv, 4, 20-38.

c11.20

⁴ iv, 3, 7-12.

c11.21

⁵ The list of the Puranas in Viṣ iii, 6, 22-4 could not have been completed until the Bhāgavata and other very late Puranas had come into existence.

c11.22

Three other Puranas contain all or nearly all the genealogies, the Garuḍa, Agni ⁶ and Bhāgavata.⁷ Their accounts are all late re-compilations, the Bhāgavata being one of the very latest, about the ninth century A.D. They do not reproduce any of the old verses except rarely, but have re-stated the genealogies in fresh verses, generally in more condensed form. The Garuḍa and Agni give merely bald pedigrees with hardly any incidental allusions. The Garuḍa and Bhāgavata follow the Vāyu tradition as regards the Aikṣvākus, and the Agni the Matsya tradition. As regards the other dynasties, all three follow in a general way the common tradition. The Bhāgavata has used the Viṣṇu in its composition, and so also has the Garuḍa apparently. All three however have peculiarities of their own. The Agni has erred seriously as regards the Kānyakubja and Kāśi dynasties. The Bhāgavata is fuller and contains stories and allusions, which show a marked brahmanical colouring and some corruptions; and it has taken considerable liberties with names. These three Puranas have no authority as regards the genealogies, yet are of use for comparison, and sometimes elucidate the older accounts. The Bhāgavata has a special value, where it restates traditions, found in the older Puranas, that do not harmonize with brahmanical assertions and pretensions and could not have originated with the brahmans; for the fact that it, a thoroughly brahmanical composition, affirms such traditions, is very strong testimony that the traditions were genuine and could not be discarded. It will be often cited therefore in this way.

c11.23

⁶ Both in Jīvānanda Vidyāsāgar's editions.

c11.24

⁷ The Gaṇapata-Kṛṣṇajī edition.

c11.25

The Liṅga account is based on the ‘Vāyu version’, but adapts it to frame its own text. Often it has the same verses, but often also it modifies, curtails, and freely omits, especially incidental and descriptive matter; and it adds occasionally.¹ It also interpolates religious teaching, as where it introduces a long eulogy of Rudra (i, 65, 46 f.). It however shows traces of the influence of the Matsya version; thus it has the same verses sometimes,² and ends the Aikṣvāku genealogy with six kings instead of the Vāyu’s twenty-one, just as the Matsya does. The Liṅga is useful for collating with the ‘Vāyu version’ when the verses agree, and for comparison where they vary; thus, it suggests (i, 65, 42-3) that the Vāyu’s misplaced lines (88, 79b, c) about the Viṣṇuvṛddhas³ should probably come after verse 74.

c11.26

¹ e. g. compare Lg i, 66, 1 f. with Vā 88, 77 f.

c11.27

² e. g. Lg i, 66, 14b-20a agree with Mat 12, 39-44.

c11.28

³ Bd iii, 63, 70-80 omit these lines. Cf. Kūr i, 20, 28.

c11.29

The Kūrma account is a composite production. Now and again it has a few lines like the Vāyu text,⁴ and like the Matsya text,⁵ in the Aikṣvāku genealogy, but it follows the Matsya rather, where they differ. It is a late composition and shows brahmanical features; thus it omits most of the Vāyu’s tales and introduces brahmanical fabrications instead: for instance, it makes Gautama (who was far later) a contemporary of Yuvanāśva I,⁶ and tells long fables about king Vasumanas⁷ and the Haihaya kings Jayadhvaja and Durjaya.⁸

c11.30

⁴ Thus Kūr i, 21, 16a and 17 agree with Vā 88, 183a and 184.

c11.31

⁵ Kūr i, 21, 4-8a are part of Mat 12, 39-44.

c11.32

⁶ i, 20, 13-18.

c11.33

⁷ i, 20, 32-76.

c11.34

⁸ i, 22, 22-80; 23, 6-44.

c11.35

The Śiva gives only the account of Manu and his offspring (vii, 60, 1), the Aikṣvākus (ibid., 33) and the Śaryātas (ibid., 20). Its text is similar to that of the ‘Harivaṃśa version’, but is less accurate and shortens or omits incidental and descriptive matter.

c11.36

The Mārkaṇḍeya gives only the account of Manu’s offspring (chaps. 111, 112) and the early part of the Vaiśala dynasty (chaps. 113–36, 109–10), and it narrates the latter at immense length with abundant imaginative description. Its text is its own.

c11.37

The accounts in the Mahābhārata¹ are peculiar. They are partly in verse and partly in prose, and do not appear to be ancient. They will be noticed in connexion with the genealogies that they treat of.

c11.38

¹ Calcutta edition.

c11.39

The Rāmāyaṇa² is, as it professes to be, altogether a brahmanical book. Some of its genealogies agree with those in the Puranas, and where it differs from them, as in the Aikṣvāku line, it is manifestly wrong, as will be shown in chapter VIII. The legends it narrates or mentions are generally distorted according to brahmanical notions and through the brahmanical lack of the historical sense.³

c11.40

² Bombay edition.

c11.41

³ e. g. its statement that Yadu was virtually a rākṣasa and his offspring were rākṣasas and yātudhānas (vii, 59, 14, 15, 20) is outrageous.

c11.42

In examining the genealogies it is of little profit and is likely to be misleading to deal with the accounts in the several Puranas separately. The only trustworthy course is first to collate the texts that generally agree and ascertain as far as possible what original text they indicate, and then construct the genealogy therefrom. By this method individual corruptions and errors can be corrected, losses and omissions remedied, and interpolations and alterations detected with reasonable confidence; and thus a text may be framed which approaches as nearly as is possible to the common original on which all those texts were based. At times divergences occur which are greater than can be so resolved, and we find competing texts, yet they are not on the whole so frequent or serious as to cause material difficulty; and small discrepancies do not really affect the general fabric of the genealogy.

c11.43

The method here advocated cannot be merely one of pure verbal criticism; some scope must be allowed to discrimination and judgement based on a general study of the Puranic texts. Whether one’s individual decisions on the reconstruction of the passages be sound or not must be tested by study devoted to the Puranas; but the principle is sound and provides the only sane way in which these genealogies can be examined. This is the method used here. The texts of the Puranas have been collated throughout, wherever doubtful points arose; and wherever it is essential to elucidate the discussion, the text that appears to be most probably the nearest approach to the original common text is quoted. It will be often found that, though there may be scope for difference in the selection of particular words in framing the text, yet that does not affect the general sense of the passages, the purport being clear, though the words selected may vary. As an illustration of the advantage of collating the texts may be given the following passage, which shows how important may be what appear to be stray readings.

c11.44

In the Solar line of Ayodhyā there were three early kings, Prasenajit, his son Yuvanāśva, and his son Māndhātṛ. The Brahmāṇḍa and Vāyu say Gaurī was Yuvanāśva’s wife—

c11.45

atyanta-dhārmikā Gaurī tasya patnī pati-vratā

c11.46

and call her son Māndhātṛ Gaurika.¹ But the Brahma, Harivaṃśa and Śiva say Gaurī was Prasenajit’s wife, thus making her Māndhātṛ’s grandmother.² The question arises, which of these accounts is right? Now there was in the Paurava line an early king whose name is given as Matināra,³ Antināra⁴ and Ratināra,⁵ and we may adopt Matināra as the form most often found, though Atināra may be the true original. The Vāyu, Matsya, and Harivaṃśa itself say he had a daughter Gaurī and she was mother of Māndhātṛ;⁶ the Harivaṃśa thus contradicting its statement in the former passage. It is clear then that she was wife of Yuvanāśva and not of Prasenajit; and the phrase atyanta-dhārmikā in the above line supplies an interesting corroboration of this. This phrase is the general reading in the Brahmāṇḍa and Vāyu, but two copies of the Vāyu read Atimānātmajā instead, which is obviously a mistake for Atinārātmajā or Matinārātmajā, as she was Matināra’s daughter; and this rare reading is the right one and confirms the statement in the Paurava genealogy. But it was corrupted and was not understood, and so was altered to the intelligible but commonplace epithet atyanta-dhārmikā, which has now almost superseded that true reading. The Brahmāṇḍa and Vāyu therefore have the correct relation, and the three other Puranas have altered it. It is possible to suggest a reason for their tampering with it. It was a very common trait to provide explanations of names, as shown in chapter V, and Māndhātṛ's name is explained in an absurd fable, which says he was born from his father Yuvanāśva's side.¹ That being accepted, it followed that Gaurī was not his mother, and so it was natural to transfer her as wife from Yuvanāśva to Prasenajit.

c11.47

¹ Bḍ iii, 63, 66–8. Vā 88, 64–7 (°dhārmiko by mistake).

c11.48

² Br 7, 90–2. Hv 12, 709–11. Śiv vii, 60, 74–6.

c11.49

³ MBh i, 94, 3703; 95, 3778. Br 13, 51. Hv 32, 1715. Ag 277, 4.

c11.50

⁴ Mat 49, 7.

c11.51

⁵ Vā 99, 128–9. Viṣ iv, 19, 2. Gar 140, 4.

c11.52

⁶ Vā 99, 130. Mat 49, 8. Hv 32, 1716. Bḍ wanting.

c11.53

¹ MBh iii, 126, 10423–53; vii, 62, 2274–5; xii, 29, 974: copied in Viṣ iv, 2, 13–18. See pp. 40 and 76.

Linked texts (18)