Skip to main content

The consecration of Kumāra: The role of Thanesar and King Harṣa in the composition of the Skandapurāṇa


In ‘The World of the Skandapurāṇa’, published in 2014, Hans T. Bakker presents a possible scenario for the composition of the Skandapurāṇa1 (SP), dating its conception to the reign of the Maukhari king Avantivarman in the late sixth century and the completion of its first recension to the times of king Harṣa in the first half of the seventh. Because of the attention it receives in the text, Vārāṇasī is identified as the most likely place of origin;2 the authors of the text belonged to the Pāśupata Śaivas, writing for a broad audience of mostly lay Māheśvaras.3 In this paper I would like to comment on some aspects of this scenario by investigating the relations between the SP and king Harṣa – both his person and his region of origin, the kingdom of Thanesar. I will argue that there are internal indications that the original constituent parts of the text may originate from there — and that its further development may have been impacted by the political reality of Harṣa having come to power. My argument will consist of three parts: the first related to the geography of the sacred places mentioned in the SP, the second concerning the way Harṣa is alluded to in the chapters describing the birth and consecration of Skanda and the third about the expansion of the original, “core” part of the SP.

Hide All

I would like to thank the members of the Skandapurāṇa team for their feedback on earlier drafts, and especially Hans Bakker and Peter Bisschop, whose comments and suggestions have greatly improved this paper. The initial research was done in the context of the project A Historical Enquiry Concerning the Composition and Spread of the Skandapurāṇa, funded by the Netherlands Organisation for Scientific Research (NWO).

Hide All

1 For an introduction to this text and the critical edition currently in progress, see Adriaensen, R., Bakker, H.T. and Isaacson, H., The Skandapurāṇa. Volume I. Adhyāyas 1–25 (Groningen, 1998).

2 Bakker, H.T., The World of the Skandapurāṇa (Leiden, 2014), pp. 138140. See also: Bakker, H.T. and Isaacson, H., The Skandapurāṇa. Volume IIA. Adhyāyas 26–31.14. The Vārāṇasī Cycle (Groningen, 2004), p. 52 n. 174.

3 For a discussion of the connections between the SP and the Pāśupata movement, see Bakker, World of the Skandapurāṇa, pp. 138–151.

4 Bisschop, P.C., Early Śaivism and the Skandapurāṇa (Groningen, 2006), p. 14.

5 SP 7.35; SP 7.24 and SP 7.36; SP 7.25.

6 Bakker, World of the Skandapurāṇa, pp. 166–169.

7 Granoff, P., ‘Mahākāla's Journey: from Gaṇa to God’, Rivista degli Studi Orientali, LXXVII (2004), pp. 100102.

8 HC Chapter 7, section 226. Bāṇa's Harṣacarita is a contemporary, but highly poetical, partial biography of Harṣa. I use Führer's text edition (Führer, A. A., Bāṇabhaṭṭa's Biography of King Harshavardhana of Sthāṇvīśvara with Śaṅkara's commentary (Bombay, 1909)), but refer to the sections as given in the translation by Cowell and Thomas: Cowell, E. B. and Thomas, F. W. (translator), The Harṣa-Carita of Bāṇa (London, 1897).

9 SP 9.23-24

10 Adriaensen, Bakker and Isaacson, Skandapurāṇa Vol. I, p. 77 n. 50.

11 SP 12.25

12 Bakker, H.T. and Entwistle, A., Vaiṣṇavism. The History of the Kṛṣṇa and Rāma Cults and their Contribution to Indian Pilgrimage (Groningen, 1981), p. 120.

13 Ibid., p. 119.

14 Ibid., p. 120.

15 SP 19.15.

16 SP 22.29.

17 Adriaensen, Bakker and Isaacson, Skandapurāṇa Vol. I, p. 101, n.134.

18 Bisschop, Early Śaivism, pp. 12–14.

19 Ibid., p. 6.

20 MBh. 3.81.14. I refer to the critical edition of the MBh: Sukthankar, V.S. et al. , The Mahābhārata (Poona, 1927–59).

21 We do not know the precise extent of the Thanesar kingdom; given the prominent place its royal family was to obtain in North Indian politics however, it must have been an appreciable size.

22 For an overview of the anukramaṇikā see Adriaensen, Bakker and Isaacson, Skandapurāṇa Vol. I, pp. 61–64; for a discussion of its contents see ibid., pp. 55–56 and Törszök, J., ‘Three Chapters of Śaiva Material Added to the Earliest Known Recension of the Skandapurāṇa’, in Origin and Growth of the Purāṇic Text Corpus, (ed.) Bakke, H.T.r (Delhi, 2004), pp. 2628. I will come back to the anukramaṇikā in the third part of this paper.

23 Adriaensen, Bakker and Isaacson, Skandapurāṇa Vol. I, p. 61.

24 This possibility is also discussed in Bakker, H.T., Bisschop, P.C., and Yokochi, Y., The Skandapurāṇa. Volume IIB. Adhyāyas 31–52. The Vāhana and Naraka Cycles (Leiden, 2014), p. 4, although there it is argued that the Vārāṇasīmāhātmya changed place as a result of the insertion of the Vāhana and Naraka cycles. If I understand the editors' argument correctly, the anukramaṇikā would give us the original place of the māhātmya. This would, however, still place it in the midst of the chapters concerning Kurukṣetra – besides, I do not see why the insertion of the Vāhana and Naraka cycles would necessitate the displacement of the māhātmya. Nonetheless, I will come back to this possibility when discussing the anukramaṇikā.

25 SP 25.57; SP 31.14

26 īpsitaṃ saha devyā vai jagāma sthānam avyayam ‖ (SP 25.57cd)

atha dṛṣṭvā tato devīṃ devadevo vṛṣadhvajaḥ |

tapyato mandarasyāśu varadānārtham abravīt ‖ (SP 31.15)

27 Bisschop, Early Śaivism, p. 192.

28 I do not agree with Bisschop's suggestion that ‘This incongruity seems to imply that Japyeśvara in SP 21–22 and SPS 62–63 refer in fact to two different places.’ (Ibid., p. 192).

29 See Bakker, World of the Skandapurāṇa, p. 79 n. 224.

30 SP 32: Bhadreśvara, Kubjāmraka, Brahmāvarta and Kanakhala; SP 33: Madhyama. For the location of the localities from SP 32, in the Hardwar region some 130 km east of Thanesar, see Bakker, World of the Skandapurāṇa, pp. 174–186. The location of Madhyama is probably to be found around Kedarnath, some 125 km northeast of Hardwar.

31 Y. Yokochi, The Skandapurāṇa. Volume III. Adhyāyas 34.1–61, 53–69. The Vindhyavāsinī Cycle (Leiden/Groningen, 2013), pp. 5–7. These layers are narrative layers and do not necessarily have any implications for the textual history of the SP. I will come back to this cycle in more detail in the second part of this paper and discuss the textual history of the SP in the third part.

32 Ibid., p. 6 n. 9.

33 SP 62. See Bakker, World of the Skandapurāṇa, pp. 221–239 for a discussion of this locality.

34 Hiltebeitel, A., ‘Conventions of the Naimiṣa Forest’, Journal of Indian Philosophy, 26.2 (1998), pp. 169170.

35 Ibid., p. 167.

36 See Ibid., pp. 167–169.

37 MBh 9.41–46. I will come back to the citation of this passage in the SP.

38 SP 4.36–39. Brahmā casts a cakra which the seers should follow. Where the rim of the cakra breaks, the seers have to hold a great sacrifice. That place is Naimiṣa.

39 In SP 1.8–9 it is mentioned that the Mahābhārata and the Purāṇa were told there by the same sūta that tells the SP.

40 Bakker, World of the Skandapurāṇa, pp. 123–128.

41 This would be the case for both Bakker's dating of the SP to the period 570–620 (ibid., p. 137) and my own, which I will discuss in the next sections.

42 For the text of SP 70 and onwards, I rely on the following edition: Kṛṣṇaprasādabhaṭṭarāī, Skandapurāṇasya Ambikākhaṇḍaḥ (Kathmandu, 1988).

43 SP 164.1

44 Beal, S., Si-Yu-Ki. Buddhist Records of the Western World. Vol. I (Delhi, 1969), pp. 211213.

45 See Deeg, M., ‘The Political Position of Xuanzang: The Didactic Creation of an Indian Dynasty in the Xiju ji’, in The Middle Kingdom and the Dharma Wheel, (ed.) Jülch, Th. (Leiden, 2016) for a discussion of Xuanzang's motives in writing for the Chinese emperor Taizong.

46 For a description of the political situation see Bakker, World of the Skandapurāṇa, pp. 77–93 and Devahuti, D., Harsha. A Political Study (Oxford, 1970), pp. 7074.

47 As Bakker has observed, also the names of these enemies are similar: Śaśāṅka (Moon) and Tāraka (Star). (Bakker, World of the Skandapurāṇa, pp. 17–18, 162.)

48 Beal, Si-Yu-Ki, p. 213.

49 SP 163.32-45. I will come back to this passage later on, discussing it in connection with a similar passage in the Harṣacarita.

50 HC Chapter 2, section 78.

51 HC Chapter 3, section 100.

52 Cowell and Thomas, Harṣa-Carita, p. 76.

53 HC Chapter 3, section 100.

54 Führer, Bāṇabhaṭṭa's Biography, p. 139. This would provide the alternative translation: “In him Skanda has made known his power/spear by laying down Tāraka at one stroke”.

55 Bakker, World of the Skandapurāṇa, pp. 156–157.

56 HC Chapter 3, Section 95

57 The immediate context makes it clear that the recitation is of a small part of a larger work (“… he turns over the leaf marking the end of the chapter read in the morning session, takes a small bundle of some folios and recites the ‘Purāṇa spoken by the Wind.’” – translation in Bakker, World of the Skandapurāṇa, p. 156). I would therefore think it possible to translate pavamānaproktaṃ purāṇaṃ somewhat freely as “the part of the purāṇa spoken by the Wind”.

58 Kane reads vaṃśānugamavivādi as one compound and translates: “where disputants follow their family”. (Kane, P.V., The Harshacarita of Bāṇabhaṭṭa. Text of Uchchhvāsas I-VIII (Bombay, 1918), p. 311.)

59 Translated by Cowell and Thomas, Harṣa-Carita, p. 73.

tad api munigītam atipṛthu tad api jagadvyāpi pāvanaṃ tad api |

harṣacaritād abhinnaṃ pratibhāti hi me purāṇam idam ‖

vaṃśānugam avivādi sphuṭakaraṇaṃ bharatamārgabhajanaguru |

śrīkaṇṭhaviniryātaṃ gītam idaṃ harṣarājyam iva |

60 There are also other ways to read śrīkaṇṭhaviniryātaṃ; Kane mentions “issuing from a sweet throat” and “issuing from Śrīkaṇṭha”, where Srīkaṇṭha can either refer to the country or to Śiva: “Then gīta would refer to the Vāyupurāṇa and not the chanting of it by Sudṛṣṭi.” (Kane, Harshacarita, p. 312.) The conclusion that the purāṇa must be from Thanesar is therefore not a necessary one, but certainly a possible reading.

If we do take these verses to refer to the SP, I would translate the second verse as follows:

“Following upon the genealogies (i.e. the Vāyupurāṇa, which is briefly summarised in SP 5.5–9), speaking of its lord” (avi-vādi), in which the deeds (of Harṣa) are clear, weighty by sharing in/honouring (bhajana) the Mahābhārata (bharatamārga),

issuing from Thanesar, this poem* is like the reign of Harṣa!”

* For a translation of this verse with meanings related to gītam as “song” (i.e. in performance), see Kane, Harshacarita, pp. 311–312 or Bakker, World of the Skandapurāṇa, pp. 155–156.

61 HC Chapter 3, Section 101. Translated by Cowell and Thomas, Harṣa-Carita, p. 76.

kasya na dvitīyamahābhārate bhaved asya carite kutūhalam

62 See also Bakker, World of the Skandapurāṇa, pp. 163–164.

65 Sharma, B. N., ‘Abhiṣeka in Indian Art’, Journal of the Oriental Institute, XXI (1971–1972), p. 110.

66 This information was made available to me by Hans Bakker.

67 The text edition I use is Smith, D., The Birth of Kumāra by Kālidāsa (New York, 2005).

68 Bakker, World of the Skandapurāṇa, p. 165 n. 506.

tārakākhyena no lakṣmīr āmṛṣṭā prasabhaṃ prabho | (SP 163.66ab)

evaṃ yad āttha bhagavan āmṛṣṭaṃ naḥ paraiḥ padam |

pratyekaṃ viniyuktātmā kathaṃ na jñāsyati prabhuḥ ‖ (KS 2.31)

69 These correspondences are also discussed in Wattelier-Bricout, A., ‘Du tableau esthétique au récit de bataille: un exemple sanskrit de transposition de genre du Kumārasaṃbhava au Skandapurāṇa’, in Actes des 19èmes Rencontres Jeunes Chercheurs en Sciences du Langage (Paris, 2017).

70 vajraṃ vakṣaḥsthale tasya kaṭhine kuṇṭhatāṃ yayau ‖

cakraṃ hareḥ sahasrāraṃ jvalitaṃ jvalitaujasaḥ |

āsādya mṛdutāṃ bheje kaṇṭhaṃ tasyāmaradviṣaḥ ‖

daṇḍo vaivasvatasyāpi pradīptānalabhāsvaraḥ |

alātam iva toyaughe nirvāṇo dānavorasi ‖

samare ripudurvāro mahāpāśaḥ pracetasaḥ |

tasya kaṇṭhe samabhavan mantreṇeva vaśīkṛtaḥ ‖ (SP 164.66c-69)

Compare (in the order of the SP):

vṛtrasya hantuḥ kuliśaṃ kuṇṭhitāśrīva lakṣyate ‖ (KS 2.20)

… pratighātotthitārciṣā |

haricakreṇa tenāsya kaṇṭhe niṣka ivārpitaḥ ‖ (KS 2.49)

yamo ‘pi … daṇḍeṇāstamitatviṣā |

kurute ‘sminn … nirvāṇālātalāghavam ‖ (KS 2.23)

kiṃ cāyam aridurvāraḥ pāṇau pāśaḥ pracetasaḥ |

mantreṇa hatavīryasya phaṇino dainyam āśritaḥ ‖ (KS 2.21)

71 See Bakker, World of the Skandapurāṇa, pp. 4–5. After all, Śiva will be able to solve the problem that Viṣṇu can not.

72 samīpe tasya sāśaṅkas tālavṛntāyate ‘nilaḥ |

svapuṣpanikarair bhītā ṛtavas tam upāsate ‖

havirbhuji hutaṃ havyam ācchinatti balād asau |

miṣatāṃ sarvadevānāṃ devadeva diteḥ sutaḥ ‖ (SP 163.71-72)


…sādhvasāt | na vāti vāyus tatpārśve tālavṛntānilādhikam ‖ (KS 2.35)

… puṣpasambhāratatparāḥ |

… ṛtavas tam upāsate ‖ (KS 2.36)

yajvabhiḥ saṃbhṛtaṃ havyaṃ vitateṣv adhvareṣu saḥ |

jātavedomukhān māyī miṣatāṃ ācchinatti naḥ ‖ (KS 2.46)

73 SP 163.33–48

74 Bakker, World of the Skandapurāṇa, p. 165.

75 Ibid., p. 85.

76 Ibid., 165. The euphemism consists of an offer and refusal of kingship and appears in both cases closely before the taking of Lakṣmī/Rājyaśrī is reported.

77 atha prekṣya bhayaṃ tasmād ātmano balasūdanaḥ | (SP 163.23ab)

78 atha darpaṃ vihāyāśu praṇeme pākaśāsanaḥ |

tejasāṃ nidhaye tasmai prāha caivaṃ vinirjitaḥ ‖

gṛhāṇa rājyam … | (SP 163.32-33a)

79 na mamendratvam īpsitam ‖ (MBh 3.218.14d)

80 anena bahudoṣeṇa (SP 163.36c)

doṣāṇāṃ nityam āvāsaṃ (SP 163.42c)

81 nāham arthī (SP 163.36d)

rājyaṃ … na jātu manasāpy aham |

… kāmaye … ‖ (SP 163.42)

82 “The government of a country is a responsible office and ever attended with difficulties.” (Beal, Si-Yu-Ki, p. 211.) In Deeg's translation: “The burden of the heir to the throne has always been hard to [carry].” (Deeg, ‘The Political Position of Xuanzang’, p. 113.)

83 HC Chapter 6, Section 201–203. See also Bakker, World of the Skandapurāṇa, p. 165.

84 It should be noted that the selection of this specific passage from the MBh in itself is significant, since the rest of the story of Skanda as told in the SP is very different from the one told in the Āraṇyakaparvan.

85 The real reason probably being political. It may be that Harṣa was suspected of having some hand in the murder of Rājyavardhana and that assuming the title of king would have strengthened those suspicions. Bakker (World of the Skandapurāṇa, p. 86 n. 248) calls the view that Harṣa may have had something to do with the death of Rājyavardhana historical speculation, but I think that it is not necessary to assume Harṣa's complicity – the suspicion alone would have been enough. Such a suspicion would have been natural; after all, Harṣa was next in line for the throne.

In Goyal, S., ‘The Recently Discovered Kurukshetra-Varanasi Grant of Harṣa: Year 23’, East and West, 57 (2007), pp. 199200 an overview is given of the arguments for Harṣa's involvement in Rājyvardhana's murder. The most interesting of these is a particular quotation from the Harṣacarita: amarapatir ivāgrajavadhakalaṅkaprakṣālanākulaḥ (Chapter 7, Section 233). That this is a reference to the killing of Rājyavardhana is confirmed by Śaṅkara: agrajo jyeṣṭho rājyavardhano dvijaś ca (Führer, Bāṇabhaṭṭa's Biography, p. 281). If we take kalaṅka to mean ‘blame’ here, we can translate: “like Indra, occupied with washing away the blame for killing his eldest brother”. This would fit very well with the immediate context: when setting out for the march against Śaśāṅka, Harṣa appears before the army; immediately before our quotation it is said: “The weight of his great qualities seemed to sink him deep in royal hearts all moist with affection, the flood of his magnificence to anoint the very marrow of the beholders” (Translation by Cowell and Thomas, Harṣa-Carita, p. 204). Harṣa's qualities and the affection they evoke publicly show the accusation to be baseless: such a man could never commit such a foul act. Interpreting kalaṅka as “stain” – i.e. “guilt” – would make far less sense: why would Bāṇa include a reference to such a crime, committed by the hero of his story?

It may be significant that in the model passage from the MBh, Indra argues that – if he were to remain king – the people would split into factions supporting either him or Skanda, which would result in war (MBh 3.218.16-17). Although this does not seem to be mentioned in the SP (Bhaṭṭarāī’s text is somewhat corrupt at this point), it may be alluded to by the very inclusion of the kingship discussion. If this reflects some historical reality, it may be that the faction that supported Rājyavardhana was still active at this point and would not allow Harṣa to be crowned king.

86 Bakker has suggested to me that it might have been inspired by the attempted rape of Devasenā by the asura Keśin in MBh 3.213, which is told as an introduction to the story of the birth of Skanda. While this story element perhaps crossed the minds of the authors of SP 163, I fail to see any real correspondence here: Devasenā is rescued by Indra, while Tāraka succeeds; the figures of Devasenā and Lakṣmī and those of Keśin and Tāraka do not have similar roles in the story. As such, I do not think this passage in the MBh alone could have inspired the change made by the SP authors to the KS verse.

87 I will come back to the issue of dating the SP later on; I will then argue for a dating of SP 163–165 prior to the HC.

88 utpāṭya prasabhaṃ tena nandane kalpapādapāḥ |

niveśitā yathākāmaṃ svagṛhodyānabhūmiṣu ‖ (SP 163.70)

89 tenāmaravadhūhastasadayālūnapallavāḥ |

abhijñāś chedapātānāṃ kriyante nandanadrumāḥ ‖ (KS 2.41)

90 utpāṭya meruśṛṅgāṇi kṣuṇṇāṇi haritāṃ khuraiḥ |

ākrīḍaparvatās tena kalpitāḥ sveṣu veśmasu ‖ (KS 2.43)

91 Although Bāṇa ascribes the murder of Grahavarman, and implicitly the imprisoning of Rājyaśrī, to the Mālava king, the attack on Kanauj probably was a joint effort of “a confederation of all those who held a grudge against the Maukharis, spearheaded by Śaśāṅka” (Bakker, World of the Skandapurāṇa, p. 91). In any case, by the time of Harṣa's consecration, the Mālava king was already defeated by Rājyavardhana and no longer a threat.

92 See e.g. KS 8.27.

93 Bāṇa refers to Lakṣmī as the sister of the Kaustubha gem, also born at the churning of the ocean: sodaryalakṣmīcumbanalobhena kaustubhamaṇer iva mukhāvayavatāṃ gatasyādharasya galatā rāgeṇa pārijātapallavaraseneva siñcantaṃ diṅmukhāni; “The redness of his lip bedewed all the regions of space like the exudation of a branch of the heavenly tree, and the lip itself seemed as if it were the Kaustubha gem, takings its place as a feature of his face in its desire to kiss its sister Lakṣmī who sat enthroned there.” (HC Chapter 2, Section 81; translation by Cowell and Thomas, Harṣa-Carita, p. 60). If we take sodaryalakṣmīcumbanalobhena to go with pārijātapallavarasena we even have a direct reference to Lakṣmī and Pārijāta being siblings.

94 kusumasaurabhagarvabhrāntabhramarakulakalakalapralāpasubhagaḥ pārijāta iva jātaḥ śriyā saha punar avatārito medinīm (HC Chapter IV Section 161; translation by Cowell and Thomas, Harṣa-Carita, p. 127).

95 prabhālāvaṇyamadasaurabhamādhuryaiḥ kaustubhaśaśimadirāpārijātāmṛtaprabhavaiḥ sarvaratnaguṇair aparām iva surāsuraruṣā ratnākareṇa kalpitāṃ śriyam (HC Chapter IV section 162; translated by Cowell and Thomas, Harṣa-Carita, p. 129.)

96 deva jāmātur antikāt tāmbūladāyakaḥ pārijātakanāmā saṃprāptaḥ (HC Chapter IV Section 160; translated by Cowell and Thomas, Harṣa-Carita, p. 126.)

97 For another perspective on this passage, see Wattelier-Bricout, ‘Du tableau esthétique’.

98 Bakker, World of the Skandapurāṇa, p. 155.

99 Ibid.,

100 Mann, R. D., The Rise of Mahāsena. The Transformation of Skanda-Kārttikeya in North India from the Kuṣāṇa to Gupta Empires (Leiden, 2012), p. 196.

101 That this is not just a trivial rearrangement of the story elements follows from the fact that another reason for Skanda's birth is provided. In layer A of the Kauśikī-Vindhyavāsinī Cycle, it is told how Pārvatī wishes for a son and obtains this boon after practicing tapas. See Yokochi, Skandapurāṇa Vol. III, pp. 22–24.

102 B.L. Staiger. The Skanda Myth in the Skandapurāṇa. A Critical Edition and Study of Adhyāyas 163 to 165 of the Skandapurāṇa. PhD thesis (Hamburg, 2017), p. 91 n. 120.

103 Ibid.

104 “a metaphor for the delegation of royal power that comes with the consecration” (Ibid., p. 50).

105 In the Mahābhārata, it is only called paramāsana, which is less specific. (MBh. 9.44.2; see Staiger, Skanda Myth, pp. 50–51).

106 “it is striking that the participation of the Lokapālas in the consecration is highlighted because it symbolises the transfer of power over the world from the deities guarding the world to Skanda.” (Ibid., p. 91 n. 120).

107 For Staiger's discussion of these passages see ibid., pp. 50–51, 55.

108 SP 165.38–44.

109 See Bakker, World of the Skandapurāṇa, pp. 97–102 for a historical description of Harṣa's campaigns.

110 citraṃ dadhānaḥ kavacaṃ mahādyutir nibaddhanistriṃśatalāṅgulitravān |

bibhrad dhanuḥ sajjam apīddhapauruṣo jahau śaśāṅkaḥ sahajāṃ na saumyatām ‖ (SP 165.8)

111 Two other curious instances of the word śaśāṅka are found in SP 164 (verse 17: śaśāṅkakiraṇatviṣam and verse 57: śaśāṅkakiraṇadyutim). Both adjectives describe a crown that is presented to Skanda (or, in the first case, if we take hāraṃ to be a noun, a necklace), and can perfectly well be translated as “shining like moonbeams”. It is remarkable, though, that in the crown, a symbol of royalty, we find as it were a reflection of Śaśāṅka, as if he already bows down and acknowledges the other ruler's sovereignty.

112 In this context two instances of the word harṣa should be mentioned. At the beginning of SP 164 the gods come to Samantapañcaka for Skanda's consecration. The first to be mentioned in the list of the assembled gods is Kṣiti, who is said to be harṣasaṃphullalocanā (v. 5, “with eyes wide of joy”). Further on Gaṅgā is mentioned, who has come there because of joy (v. 12, ājagāma tadā harṣāt). In both cases it would be possible to read harṣa as Harṣa in a meaningful way, but since harṣa is a common word we cannot draw any real conclusions from this. Still, it is worthy of mention that in exactly those passages that have no parallel in the MBh model (see our earlier discussion of SP 164), both the words harṣa and śaśāṅka are mentioned twice (see previous note).

113 Y. Yokochi, The Rise of the Warrior Goddess in Ancient India. A Study of the Myth Cycle of Kauśikī-Vindhyavāsinī in the Skandapurāṇa. PhD thesis (Groningen, 2004), pp. 47–49.

114 Ibid., p. 123 n. 105. I can add that there are also textual correspondences between Kauśikī’s battle with Sumbha and Nisumbha in SP 64–66 and Skanda's battle with Tāraka in SP 165. SP 64.1–5, 9 is very similar to SP 165.1-5, the catalogue of bad omens in SP 64.12-14 is similar to the one in SP 165.13-16, SP 65.3 shows a similar idea to SP 165.29, SP 66.9 and 13 are clearly inspired by SP 165.27, SP 66.10c is identical to SP 165.26c. There may be more.

115 Ibid., pp. 122–123.

116 SP 164 is heavily based on MBh. 9.44–45. See Adriaensen, Bakker and Isaacson, Skandapurāṇa Vol. I, p. 26 n. 103 and Yokochi, Rise of the Warrior Goddess, p. 100 n. 59 and p. 101 n. 62 for details. The parallel is discussed in detail in Staiger, Skanda Myth, pp. 55–68.

117 Bakker, conversely, has suggested to me that it might be a sign of haste on the part of the authors.

Staiger argues throughout his thesis that changes to the literary models (Mahābhārata, Rāmāyaṇa, Kumārasaṃbhava) of the Skanda chapters were made at least partly in order to conceal the connection between the SP and the original texts (e.g. Staiger, Skanda Myth, p. 67). This seems unlikely to me, as the small changes made to the long lists in SP 164 certainly would not hide their provenance in any way. On the contrary, I think the literary allusions were very much deliberate, in order to place Harṣa in the context of the elevated and celebrated literary world to which the SP connects itself. It should not be forgotten that Harṣa himself was a man of letters – Bakker (World of the Skandapurāṇa, p. 124) also stresses this point – and would no doubt be receptive to this kind of literary play.

118 Except for a part of SP 72, to which I will turn shortly. It would not yet have included such larger cycles as the one about Andhaka, the length of which could have eclipsed the Skanda story. I will elaborate on this in the next section.

119 I am thinking primarily of SP 163–165. There is some indication that SP 165 at least was written before (parts of) the Goddess chapters. Compare:

chinnāni petur niśitaiḥ kṣurapraiḥ sakuṇḍalāny ānanapaṅkajāni | (SP165.27ab) with:

chinnāni petur niśitaiḥ kṣurapraiḥ samucchritāny ātapavāraṇāni |

citrāḥ patākā vividhā dhvajāś ca paraspareṇāhanane gajebhyaḥ ‖ (SP 66.9) and:

chinnāni vaktrāṇi sakuṇḍalāni bhujāḥ saśastrāś ca sucandanāktāḥ |

petur gajebhyo ‘surayūthapānāṃ devībhir ājau niśitaiḥ kṣurapraiḥ ‖ (SP 66.13),

where the two verses from SP 66 very much seem like two variations on the verse from SP 165. See also note 114.

On the other hand, SP 164 seems to refer back to SP 64, adding Kauśikī as the giver of the Mātṛs and adding some names of Goddesses who originate from Kauśikī’s body in SP 64 or are mentioned in SP 68 to the list that SP 164 borrows from the Mahābhārata. This is originally discussed by Yokochi (Rise of the Warrior Goddess, pp. 99–101, see especially note 61) and taken up again by Staiger (Skanda Myth, pp. 65–66). Staiger concludes “that the authors of the Skanda myth deliberately changed elements of the Mātṛ list, which they had borrowed from the Mahābhārata, in order to emphasise the connection to the Kauśikī cycle” (Ibid., p. 66). I agree, but I do not think that this proves the primacy of the Kauśikī chapters (Staiger does not argue for this, but I think the point is worth discussing). The reference might either be to the Kauśikī story in general, which would have been important and well known in the environment in which the Skanda chapters were written, or to a planned, but not yet realized, account of the story.

Staiger (Ibid., pp. 75–77) also discusses the relationship between SP 64–66 and SP 165. He notes the parallels between SP 66.10 and SP 165.26 and between SP 66.9 and SP 165.27, but seems to have overlooked SP 66.13. As a result, he claims that the two passages most probably were “composed by the same redaction group, possibly by the same author” (Ibid., p. 77). I certainly do not dispute this possibility, but I think it is important to note that, if the passages were written by the same authors, it is likely that they wrote the Tāraka chapter first.

120 There is also the possibility that work on the text had already begun earlier, the new historical events being incorporated in it by the authors.

121 See e.g. Bühler, G., ‘The Madhuban Copper-plate of Harsha, dated Samvat 25’, in Epigraphia Indica. Volume I, ed. Burgess, Jas. (Calcutta 1892), pp. 6775.

122 For a discussion of these chapters (SP 4–7) see Bakker, World of the Skandapurāṇa, pp. 165–169.

123 SP 7.13

124 athājagmuḥ surāḥ sarve sārdhaṃ tatra havirbhujā ‖ (SP 72.142cd)

125 tejo na śaknumaḥ soḍhum idaṃ tribhuvaneśvara ‖ (SP 72.144cd)

126 andhakaś cāsuro ‘tyarthaṃ tapyate duścaraṃ tapaḥ |

bhayañ ca nas tato deva dṛśyate sumahad vibho ‖ (SP 72.145c-f)

127 mā bhīr bhavatu vo devā haniṣyāmy aham andhakam |

yuṣmadarthe parākrāntaṃ hiraṇyākṣasutaṃ varam ‖

yūyañ ca merumadhyasthaṃ vanaṃ yāta suvistṛtam |

nāmnā śaravaṇaṃ nāma puṇyaṃ dvādaśayojanam ‖

yuṣmākaṃ tatra saṃbhidya kukṣīn daivatapuṅgavāḥ |

madīyaṃ tat paraṃ tejo nirgamiṣyati dīptamat ‖ (SP 72.146–148)

128 śārvaṃ tejaḥ samādāya vahniḥ kiṃ kṛtavāṃs tadā | (SP 163.1)

129 Mann does not even consider the last part of SP 72 in his summary of the Skanda story in the SP and only mentions that “… the narrative transitions to the story of Andhaka and his destruction.” (Mann, Rise of Mahāsena, p. 190.)

130 Staiger, although at first coming to a similar conslusion, now has a different view on SP 72, considering it a later addition to a core Skandapurāṇa consisting of the Kauśikī cycle (i.e. SP 60–68; see Staiger, Skanda Myth, p. 77; this roughly corresponds to Yokochi's layer B – although at other times Staiger seems to mean all the Goddess chapters, e.g. on pp. 65 and 213) and SP 163–165 (Ibid., pp. 77–81). He points to the contradiction that in SP 72 a boon is granted to Agni “that he will be able to carry and emit the semen with ease (…). But in SP 163 the semen causes Agni pain, burning him so much that he asks Gaṅgā to carry it instead” (Ibid., p. 78). While I admit that this is somewhat strange, I think that Staiger's explanation creates more problems than it solves. A core consisting only of the Kauśikī cycle and SP 163–165 does not have a continuous and cohesive narrative; in order to obtain this one would at least need layer A of the Goddess chapters (which connects these chapters with the Skanda cycle; see Yokochi, Skandapurāṇa Vol. III, pp. 22–24) and, indeed, something like SP 72. Further, Staiger's own analysis shows that SP 72, 163 and 164 are all heavily modelled on different versions of the Skanda story in the epics. The fact that SP 72 was based more heavily on the Rāmāyaṇa and the Anuśāsanaparvan, while SP 163 and 164 are based on the Āraṇyakaparvan and the Śalyaparvan does not suggest to me, as it does to Staiger, that these chapters were written by different authors, but can rather be explained by the character of the epic models themselves (As Staiger himself notes: “The content of adhyāya 72 has, however, no parallel in these parvans.” (i.e. the Āraṇyakaparvan and the Śalyaparvan; Staiger, Skanda Myth, p. 26)). In fact, the fact that SP 72, 163 and 164 were all closely modelled on the epics, as Staiger convincingly demonstrates, is a strong indication that they were indeed written by the same authors, using the same kind of writing procedure in all these chapters.

131 Yokochi, Skandapurāṇa Vol. III, pp. 15–22.

132 See e.g.: “Later, while composing chapter 55, …” (Ibid., p. 21.)

133 H.T. Bakker, P.C. Bisschop and Y. Yokochi, The Skandapurāṇa. Volume IIB. Adhyāyas 31–52. The Vāhana and Naraka Cycles (Leiden, 2014), pp. 9–10.

134 It is not, as Yokochi claims (Skandapurāṇa Vol. III, p. 16 n. 30), the beginning of the Naraka cycle, but rather, as the editors note, a shorter version of the same story (Bakker, Bisschop and Yokochi, Skandapurāṇa Vol. IIB, p. 9).

135 Hiraṇyakaśipu and Hiraṇyākṣa are brothers and the fathers of Prahlāda and Andhaka respectively.

136 Granoff, P., ‘Saving the Saviour. Śiva and the Vaiṣṇava Avatāras in the Early Skandapurāṇa’, in Origin and Growth of the Purāṇic Text Corpus, (ed.) Bakker, H.T. (Delhi, 2004), p. 129.

na śrutena na vīryeṇa na śauryeṇa na tejasā ‖

na buddhyā na ca dharmeṇa prahlādasadṛśo ‘suraḥ |

hiraṇyakaśipur nāsīd rājā tatsadṛśo mahān ‖

nāndhako na hiraṇyākṣaḥ prahlādasadṛśo bale |

bhṛguśāpāt kilāyāto daityatvaṃ viṣṇur eva saḥ ‖ (SP 172.49c-51)

137 SP 69 belongs to the Vindhyavāsinī cycle, SP 72 concerns the birth of Skanda.

138 vyāsa uvāca |

bhagavan sarvayogajña śrutam ākhyānam uttamam |

parataḥ śrotum icchāmi yad vṛttaṃ munisattama ‖

apanīya tadā viṣṇoḥ siṃharūpaṃ vṛṣadhvajaḥ |

kim anyad akarod dhīmān etad icchāmi veditum ‖

sanatkumāra uvāca |

svayoniṃ viṣṇum ānīya devadevo bhavas tadā | (SP 72.1-3b)

139 mandare sahito devyā vicacāra mahādyutiḥ ‖ (SP 72.3cd)

140 Granoff, ‘Saving the Saviour’, p. 123.

141 divo rājyaṃ punaḥ śakro lebhe niṣkaṇṭhakaṃ mahat |

divaukasaś ca sarve te svāni rājyāni bhejire ‖ (SP 71.47)

142 athāgatya tato devaḥ śūlapāṇir vṛṣadhvajaḥ |

surair vijñāpito vyāsa yat te kathitavān aham ‖ (SP 71.48)

143 The two verses read as follows:

athāgatya tato devaḥ śūlapāṇir vṛṣadhvajaḥ |

surair vijñāpito vyāsa yat te kathitavān aham

viṣṇos tyājayituṃ rūpaṃ saiṃham adbhūtakarmaṇaḥ |

śarabhas sa tadā bhūtvā himavacchikharopamaḥ ‖ (SP 71.48-49)

Alternatively, SP 71.49ab can also have been part of the original verse.

144 SP 70.3, SP 70.6

145 SP 72.23c-24

146 SP 69.9-18

There are links in vocabulary in the three door guard passages. In both SP 69 and SP 72 Nandin is described as dīptapaṭṭiśadhāriṇam (SP 69.9 and SP 72.66). In all three passages Nandin is standing at the dvāramūla (SP 69.9, SP 70.8 and SP 72.25). Both words occur nowhere else in the SP (or in the epics). The usage of dvāramūla in SP 70, however, is somewhat strange. First Nandin is said to be standing at the puradvāra (SP 70.6). Then, after hearing the gods’ request he moves (samāgatya) to the dvāramūla in order to inform Śiva.

atha nandī samāgatya dvāramūle samāsthitaḥ |

śambhor nivedayāmāsa … ‖ (SP 70.8)

The dvāramūla seems to be another place (somewhere inside) than the puradvāra. This seems to be inconsistent with SP 69 and SP 72, where the dvāramūla is where Nandin stands guard. I interpret this evidence as supporting my theory that SP 69 and SP 72 belong to the same layer, while SP 70 – the odd one out – is secondary.

147 By Nandin to Devī in SP 69.11-14, by Viṣṇu to Śiva in SP 71.54-64, by Agni to Nandin in SP 72.69-74 and by the gods to Śiva in SP 72.94-98. It should be noted, however, that such stutis are quite common in the SP.

148 See Bakker, Bisschop and Yokochi, Skandapurāṇa Vol. IIB, p. 8.

149 Pal, P., Vaiṣṇava Iconology in Nepal. A Study in Art and Religion (Calcutta, 1970), p. 5. For a discussion of this inscription see also Bakker, World of the Skandapurāṇa, pp. 245–247.

150 Pal, Vaiṣṇava Iconology, p. 4.

151 See also Granoff, ‘Saving the Saviour’, pp. 135–138.

152 surarājyamadāviṣṭaḥ śakraḥ papraccha ṣaṇmukham ‖

kas tvaṃ divyavapur bāla meruparvatam āśritaḥ |

na māṃ pūjayase dṛṣṭvā devarājam ihāgatam ‖ (SP 171.5c-6)

153 SP 163.32-45

154 See Harimoto, K., ‘Some Observations on the Revā- and the Ambikākhaṇḍa Recensions of the Skandapurāṇa’, in Origin and Growth of the Purāṇic Text Corpus, (ed.) Bakker, H.T. (Delhi, 2004), pp. 6264 for an overview of their contents.

155 Vyāsa says to Sanatkumāra: “I have now heard this whole great account of Skanda's deeds.”

156 tasmin hate daityavṛṣendrasiṃhe purā hiraṇyākṣasute pravīre |

bhagneṣu daityeṣu sadānaveṣu … ‖ (SP 168.5)

157 … tārakasya sutās trayaḥ |

tārākṣaḥ kamalākṣaś ca vidyunmālī ca pārthiva ‖ (MBh 8.24.4)

Note the slightly different forms of the names here. Instead of Kamalākṣa the SP has Maya.

158 In fact, the earlier cited SP 168.5 uses the defeat of Andhaka as the cause of the tapas of the three asuras in the same manner as the MBh uses the defeat of Tāraka (MBh 8.24.3-5).

159 kasmāt kruddho nagendraṃ taṃ krauñcaṃ dāritavān guhaḥ |

pakṣān sarvagirīndrāṇāṃ kasmāc ciccheda vāsavaḥ ‖ (SP 171.2)

The stories are divided by the Koṭīvarṣa Māhātmya (see the edition by Yokochi in Bakker, World of the Skandapurāṇa, pp. 263–264); this does not seem connected to the surrounding stories.

160 bhagavan yat tvayā proktaṃ dakṣayajñavadhe tadā |

prāpya pāśupataṃ yogaṃ brahmādyās tridivaukasaḥ ‖ (SP 174.1)

161 This was pointed out to me by Peter Bisschop.

162 evaṃ sanatkumāras tu pṛṣṭo vyāsena dhīmatā |

munīndraḥ kathayāmāsa purāṇaṃ skandasambhavam ‖

sarvāgamasamāyuktaṃ manvantarajagatsthitim |

śivayogodbhavaṃ dhyānaṃ sarvajñānārṇavaṃ mahat ‖ (SP 183.63-64)

163 Note the similarity between this word and the title of Kālidāsa's famous kāvya.

164 yaiś ca dṛṣṭāni viprendra śivaliṅgāni tāni tu |

kedāro madhyamaś caiva girīśaś ca mahālaye

catuḥśṛṅgeśvaro devaḥ śrīgirau ca maheśvaraḥ |

kārohaṇo ca deveśo gaṅgādvāre ca śaṅkaraḥ |

bhadreśaś ca mahādevaḥ śaṅkukarṇeśvaras tathā ‖

mahākāleśvaraś caiva gokarṇe ca tathā hy ubhau |

avimukteśvaraś caiva yāni cānyāni bhūtale ‖ (SP 183.47-49)

Compare with:

kedāre caiva yal liṅgaṃ yac ca liṅgaṃ mahālaye |

madhyameśvarasaṃsthaṃ ca tathā paśupatīśvaram ‖

śaṅkukarṇeśvaracaiva gokarṇe ca tathā hy ubhau |

drimicaṇḍeśvaraṃ caiva bhadreśvara tathaiva ca ‖

sthāneśvaram athaikāmraṃ kāleśvaram ajeśvaram |

bhairaveśvaram īśānaṃ tathā kārohaṇāsthitam ‖

yāni cānyāni puṇyāni sthānāni mama bhūtale |

tāni sarvāṇy aśeṣeṇa kāśipuryāṃ viśanti mām ‖ (SP 29.82–85)

165 See note 2.

166 One of these purposes would have been the assertion of Śaivism in competition with Vaiṣṇavism, a concern most clearly recognizable in the Andhaka cycle and Vārāṇasīmāhātmya. See Bakker, World of the Skandapurāṇa, pp. 4–5 and Granoff, ‘Saving the Saviour’.

167 Cecil, E.A., ‘Mapping the Pas̄úpata Landscape. Narrative, Tradition, and the Geographic Imaginary’, The Journal of Hindu Studies (2018), p. 13.

168 Translation ibid., p. 7.

brahmacārī caturthas tu kuruṣv eva sugotrajaḥ ‖ (SP 123cd)

kanyakubje tataś cānyam anugṛhya jagatpatiḥ | (SP 129ab)

169 See above.

170 Since Harṣa no doubt was the most famous person of all in North India in the first half of the 7th century, the similarities must have been quite conspicuous.

171 Lāguḍi's instructions to his disciples would hardly be applicable to Harṣa: “Free from worldly concerns, dwell along the pure banks of rivers, in sacred shrines, in abandoned houses, and in the wilderness.” (SP 167.131; translation Cecil, ‘Mapping the Pāśupata Landscape’, p. 7.)

nadītīreṣu medhyeṣu puṇyeṣv āyataneṣu ca |

śūnyāgāreṣv araṇyeṣu vāso vaḥ saṅgavarjitaḥ ‖

172 It should be noted that SP 72 is described in some detail (SP 2.16c-17b).

173 The idea of SP 2 as a blueprint was already introduced by Törszök (‘Three Chapters of Śaiva Material’, pp. 27–28), but then as a blueprint for the whole text.

174 If my reconstruction is correct, this would mean that the newly added anukramaṇikā was only used as a general guideline right from the start. An alternative theory might be that the Vārāṇasīmāhātmya was inserted at the place mentioned in the anukramaṇikā at first, but later moved in order to provide a thematic connection with the Thanesar māhātmya. Such a move was already proposed by the editors; see Bakker, Bisschop and Yokochi, Skandapurāṇa Vol. IIB, p. 4. See also note 24 above.

175 Although the Pāśupata Yoga cycle is not mentioned in the anukramaṇikā, the A recension provides two extra pādas that refer to it. Since the S recension in this chapter is only based on one manuscript, these pādas might very well be original. See Adriaensen, Bakker and Isaacson, Skandapurāṇa Vol. I, pp. 55–56.

176 Yokochi, Skandapurāṇa Vol. III, pp. 54–57.

177 Bakker, World of the Skandapurāṇa, p. 137.

178 Ibid, pp. 138–140.

179 Bakker, World of the Skandapurāṇa, pp. 4–5; see also Granoff, ‘Saving the Saviour’.

180 A reconstruction of such an event, including the recitation of the SP, can be found in Bakker, World of the Skandapurāṇa, pp. 18–20.

181 For that, the reader is referred to the critical edition of the SP, which contains chapter-by-chapter summaries. For the chapters after SP 69, an overview of the chapters’ colophons can be found in K. Harimoto, ‘Appendix’, in Origin and Growth of the Purāṇic Text Corpus, (ed.) H.T. Bakker (Delhi, 2004), pp. 139–190.

182 See Bakker, Bisschop and Yokochi, Skandapurāṇa Vol. IIB, pp. 11–12.

183 P.K. Agrawala. Skanda-Kārttikeya. A Study in the History and Development (Varanasi, 1967), Plate XVIIIa.

184 Sharma, ‘Abhiṣeka in Indian Art’, Plate II.

185 Photograph courtesy Hans Bakker.

I would like to thank the members of the Skandapurāṇa team for their feedback on earlier drafts, and especially Hans Bakker and Peter Bisschop, whose comments and suggestions have greatly improved this paper. The initial research was done in the context of the project A Historical Enquiry Concerning the Composition and Spread of the Skandapurāṇa, funded by the Netherlands Organisation for Scientific Research (NWO).

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

Journal of the Royal Asiatic Society
  • ISSN: 1356-1863
  • EISSN: 1474-0591
  • URL: /core/journals/journal-of-the-royal-asiatic-society
Please enter your name
Please enter a valid email address
Who would you like to send this to? *


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed