Lord Skanda-Murugan

'Murukan' in the Indus Script

Iravatham Mahadevan

Text of Iravatham Mahadevan's paper presented at the First International Conference Seminar on Skanda-Murukan in Chennai, Dec. 28-30, 1998.

Fig. 3: Bangles and Squirrel Sign from 'The Squirrel Sign: Title of Murukan'.
Bangles and Squirrel Sign from 'The Squirrel Sign: Title of Murukan'. 'Bangles + squirrel' (Fig. III b) is interpreted by scholars as an ancient divine title. I. Mahadevan argues with Asko Parpola that the term 'muruku' meaning 'bangles' can be traced to prehistoric inscriptions of the Indus Valley civilization.


  1. Ideograms for 'Deity' in the Indus Script
  2. Survival of the Harappan Skeletal Deity in Later Mythology and Art Traditions
  3. Identification of the Harappan Skeletal Deity with Dr. *Muruku
  4. Notes and References
  5. Bibliography

There is clear pictorial evidence from seals, sealings and other inscribed objects for the practice of religion by the Harappans1. The question whether any deity is prominently mentioned in their writing is sought to be answered in this paper.

Section I: Ideograms for 'Deity' in the Indus Script

1.1 The search for the possible occurrence of the name of a deity in the Indus Script has to be based on the following criteria :

(a) A deity conceived to be human in form (as seen in the pictorial representations) is more likely to be depicted by an anthropomorphic ideogram than by syllabic writing.

(b) The ideogram will occur with high frequency, and with especially higher relative frequency in dedicatory inscriptions on votive objects found in religious contexts.

(c) The ideogram is likely to occur repetitively as part of fixed formulas possibly representing religious incantations.

1.2 Signs 1-48 in the Indus Script are classified as ‘anthropomorphic' on the basis of their iconography 2. There are two near-identical signs in this group, Nos. 47 and 48 (Fig.1) depicting seated personages reminiscent of very similar representations of deities in the Egyptian hieroglyphic script, in which a seated figure functions as the determinative for ‘god' (Fig.2), and similar ideograms, modified by the addition of distinctive attributes, represent specific deities3. On the basis of this analogy from a contemporary ideographic script, we may assume, as a working hypothesis to begin with, that Sign 47 of the Indus Script is the ideogram for ‘deity' and that Sign 48, its modified form occurring with a much higher frequency, represents a particular ‘Deity'characterised by the distinctive attribute added to the basic sign.4 This identification receives some support from the pairing of these two signs in either order in the texts, probably to be read as ‘the deity X' or ‘X, the deity'.5

1.3 The miniature tablets and sealings found at Harappa, especially from the lower (earlier) levels, are generally considered to be votive objects with dedicatory inscriptions incised or impressed on them.6 Sign 48, one of the more frequent signs in the Indus Script, occurs with a much higher relative frequency on the votive tablets and sealings.7 Again, a text of three signs with Sign 48 in the lead, which has the highest frequency of any 3-sign sequences in the whole of the Corpus of Indus Texts, occurs almost exclusively on the votive tablets and sealings, indicating that it is a ‘religious formula' of some kind (Fig.3).8 It is significant that in the Late Harappan Period at Kalibangan, the basic ideogram for ‘deity' begins to appear as large-sized graffiti on pottery suggestive of its use also as a religious symbol (Fig.4).9

1.4 It is even more significant that the basic Indus ideogram for ‘deity' survived as a religious symbol in the Post-Harappan Era and occurs in regions far removed from the Harappan homeland:

(a) The frequent 3-sign text mentioned earlier (but with Sign 47 in the lead) is engraved on a seal found in the excavations at Vaisali, Bihar (Fig.5).10

(b) The basic Indus ideogram for ‘deity' occurs often, presumably as a religious symbol, in the pottery graffiti from the Megalithic burials at Sanur in Tamilnadu (Fig.6).11

1.5 There is thus strong prima facie evidence from iconography, context of occurrence, frequency- distribution statistics and later survivals that Sign 48 of the Indus Script represents a popular anthropomorphic deity of the Harappan Civilization. The survival of the basic Indus ideogram as a religious symbol in later times suggests that the cult of the Harappan deity spread to Eastern and Southern India along with the migration of the descendants of the Harappans to these regions after the demise of the mature Indus Civilization.

1.6 The two defining characteristics of the Harappan deity in Sign 48 are:

(a) A skeletal body with a prominent row of ribs;

(b) The deity is seated on his haunches, body bent and contracted, with lower limbs folded and knees drawn up.

1.7 As the ideogram is a conventional ‘stick figure' with no width, the side-view of the seated deity (facing left in seal-impressions) gives the appearance of ribs ‘sticking out of the body'. The Egyptian determinatives or ideograms for ‘backbone and ribs' look similar (Fig.7).12 There are also two crucial pieces of evidence, both from Kalibangan, pointing to the true nature of the ideogram:

(a) An exceptional variant of Sign 48 is found deeply incised (pre-firing) on the concave inner surface of a shallow terracotta dish (Fig.8).13 This variant depicts the deity with a large head and the backbone with four ribs ‘inside the body'.

(b) A unique seal, probably Late Harappan, found on the surface at Kalibangan, depicts a seated skeletal deity occupying the entire field (Fig.9).14 This pictorial representation may thus be classified as the ‘field symbol' equivalent of Sign 48. The deity is facing right (in the original seal), leaning forward. He has a large head and a massive jaw jutting forward. The complete ribcage is shown in clear detail with almost all the ribs in position, curving naturalistically on either side of the backbone. The deity appears to be holding a ladle (?) in his right hand. His knees are drawn up and he seems to be squatting on his haunches.15

1.8 A careful comparative study of the two crucial variant forms of Sign 48 from Kalibangan with other known variants shows that the sign is a conventional depiction of a seated skeletal figure, and that the distinctive attribute of the ‘Deity' (Sign 48) differentiating it from the ‘deity' (Sign 47) is the row of ‘ribs' (Fig.10).16

1.9 The skeletal figure appears to be a symbolic representation of the dead, or rather, the spirit of the dead, or the manes (souls of the ‘Fathers') or a demonic deity, suggesting some form of ancestor-worship.

cf. Skt. bhuta(lit., ‘who was'): a spirit, the ghost of a deceased person, a demon, imp, goblin.

preta (lit., 'the departed'): the spirit of a dead person (especially before the obsequial rites are performed), a ghost, an evil being.

Pali peta: dead, departed, the departed spirit; the Buddhist peta signifies both the manes as well as the ghosts.

Pkt. pe(y)a: a class of gods, the dead.

Ta. pey: devil, goblin, fiend. (DEDR 4438)17

1.10 The second characteristic shared by Signs 47 & 48, of being seated, denotes dignity or divinity (as in the Egyptian ideograms). The sitting posture has close parallels from the anthropomorphic sculptures found at Mohenjodaro (Pl.I).18 The bent, contracted posture serves as a linguistic clue which will be discussed in Section III.


Section II: Survival of the Harappan Skeletal Deity in Later Mythology and Art Traditions.

The identification of the ‘Harappan Skeletal Deity' leads directly to the recognition of its evolution as the ‘Emaciated Ascetics' in later Indian mythology and art traditions. Some characteristic examples are considered here.

2.1 Dadhyanca's ribs

Dadhyanca (Dadhica) is mentioned as a divinity in the Rgveda and as a teacher or rshi in the later Vedic literature and the Mahābhārata.19 Two famous myths associated with him are relevant to our study:

(a) Dadhyanca's gift of his own ribs or bones to the gods for making the vajra with which Indra slew ninety-nine Vrtras.

(b) Dadhyanca getting a horse's head by the power of the Asvins. His name and his horse-head connect Dadhyanca with Dadhikra (van), the famous divine steed presented by Mitra-Varuna to the Purus. The etymology of the two names seemingly derived from dadhi ‘curds, buttermilk' has remained inexplicable.

The myths appear to have evolved from the iconography of the Harappan Skeletal Deity remembered as a religious symbol long after its linguistic context was forgotten:

(a) ‘ribs': Dadhyanca's inseperable identification with ‘ribs and bones' suggests that he had a ‘skeletal' body.

(b) ‘horse-head': This myth must have arisen when the symbol of the Harappan Skeletal Deity was later re-interpreted as a ‘horse' with a large ‘head', four ‘legs' (though the actual number varied) and a ‘raised tail'. This interpretation is seemingly plausible when the symbol is viewed in the horizontal position. It is interesting that some modern scholars studying the Indus Script have also interpreted Sign 48 as a ‘horse' (Meriggi: ‘horse'; Misra: ‘Dadhikravan').20 The Soviet scholars have also interpreted the sign as a quadruped, but as the‘buffalo'(presumably because there is no place for the ‘horse' in their theory of the Dravidian origin of the Indus Civilization!)21

(c) The reason why Dadhyanca and Dadhikravan have names apparently derived from dadhi ‘curds' may be explained on the basis of Dravidian etymology, assuming that these are loan-translations:

muci (Ta.): to grow thin, to be emaciated (DEDR 4903).

mucar, mor (Ta.): curds, buttermilk (DEDR 4902).

murutu, muruntu (Ka.): to shrink, shrivel (DEDR 4972).

morata, morana (Skt.): sour buttermilk (connected to Dr. mucar, mor in DEDR 4902).

The Dr. words for ‘emaciated' and ‘curds' were homonymous. The Skt. names Dadhyanca and Dadhikravan appear to be the result of translating the wrong homophone, and thus ‘the emaciated one' became ‘one fond of curds'!

2.2 Bhrngin, the ‘Skeleton Demon'

Among the circle of the bhutaganas attending on Śiva, Bhrngin, the ‘Skeleton Demon', considered to be a form of Andhaka, stands out. Bhrngin got a skeletal body because of Parvati's curse when he insisted on worshipping Śiva alone and not her. Several sculptural representations of Bhrngin are known, depicting him as a mere skeleton (Pl.II).22 His antiquity, identity as a bhuta and his skeletal body indicate the derivation of the myth ultimately from the Harappan Skeletal Deity.

2.3 Bhishma and his ‘bed of arrows'

The story of Bhishma, the great pitamaha of the Kurus, is too well-known to be re-told here. Three legends connected with his deathbed as narrated in the Bhishma-vadha-parvan of the Mahābhārata are relevant to the present study:23

(a) When Bhishma fell in battle, he lay on ‘a bed of arrows' without touching the earth.

If one views the symbol of the Harappan Skeletal Deity (Sign 48) in a horizontal position, it can be interpreted as a person lying on a ‘bed of arrows' without touching the earth.

(b) When Bhishma's head was hanging down, he asked Arjuna for a pillow. Thereupon Arjuna supported Bhishma's head with three arrows shot from his Gandiva.

One of the variant forms of the Harappan Skeletal Deity (Sign 48) in which three projecting lines are seen attached to the back of the head provides the pictorial basis for this myth (See the first sign in the second row in Fig.10).

(c) When Bhishma was lying on his bed of arrows, he asked for water. Arjuna shot an arrow from his Gandiva piericing the earth, and there arose a jet of pure and cold water for Bhishma to drink.

It is interesting that the nearest pictorial depiction of this legend is provided by an Egyptian ideogram (when viewed horizontally) of ‘a man receiving purification from a stream of water' (Fig. 11).24 Perhaps a similar variant form of Sign 48 exists and may still be found.

2.4 Buddha as an ‘Emaciated Ascetic'

Gautama in the course of his wanderings in search of Truth came to Uruvela and practised the severest austerities which reduced him to a mere skeleton; but, failing to attain the goal by mortification of the flesh, he decided to take nourishment just enough to sustain the body. This famous incident in the Buddha's life is splendidly portrayed in a sculpture from Gandhara dated ca. 2-3 cent. AD. (Pl. III)25 According to tradition, the skeletal figure of the Buddha is intended as a warning to others of the futility of excessive austerities. However it is possible to take a more positive view of the depiction of the Buddha as an ‘Emaciated Ascetic in penance' as worthy of adoration. This explanation accounts in a more satisfactory manner for the wide prevalence of the motif of the ‘emaciated ascetics' in BrahmĀnical and Buddhist traditions, ultimately going back to the Harappan prototype.

2.5 Karaikkal Ammaiyar, the pey

Karaikkal Ammaiyar, the earliest of the Tamil Saivaite saints (ca. 5-6 cent. AD.), chose to describe herself in her poems as the pey, which meant originally ‘the departed soul' (from Pkt. peya), but later acquired the pejorative meanings ‘demoness, she-devil'.26 True to her assumed title, she describes Śiva's dance surrounded by ghosts; she views the ghosts as ‘blessed with sympathetic and human hearts'.27 The magnificent Chola bronzes from a later period depict her literally as the pey with a skeletal body, prominent ribcage and squatting on her haunches (Pl.IV).28 The similarity between the Gandharan Buddha and the Chola bronzes of Karaikkal Ammaiyar in the treatment of the emaciated, skeletal body is striking, even though they are wide apart in space and time. This thematic unity spanning the sub-continent and between the Indo-Aryan and Dravidian traditions indicates a common inheritance going back to the Harappan times.

2.6 The Emaciated Ascetics from Harwan

Harwan, near Srinagar in the Kashmir valley, is famous as the site where Kanishka is said to have convened the Fourth Buddhist Council in the 2nd cent. AD. The chaityagrha at this site is embellished with stamped terracotta friezes in the Late Gandharan style (ca. 4-5 cent.AD.). The most prominent among them are the repetitive friezes depicting ‘emaciated ascetics' "who are lean, nude, reduced to a skeleton, shown with their bent backs, legs tucked up, hands placed on knees and with chins resting on their hands"29 (Pl. V).30 Here too, the interpretation that the figures are intended as a warning against excessive austerities is unconvincing, especially when this depiction is the dominant motif at the site. It is more likely that the figures represent the Buddha as the emaciated ascetic. The similarity between the emaciated ascetics of Harwan and the Harappan Skeletal Deity is too close to be missed.

2.7 The Emaciated Ascetics from Paharpur

The Somapura Mahāvihara at Paharpur, Bangladesh, dating from 8th cent. AD. is especially famous for the continuous friezes comprising thousands of stamped terracotta plaques adorning the exterior walls of the plinth and the lower terraces. The plaques are known for their ‘exuberant treatment' of ‘all conceivable subjects of human interest' including divine figures, both BrahmĀnical and Buddhist (many more of the former than the latter). "Ascetics as travelling mendicants, with long beards, their bodies bent and sometimes reduced to skeletons, carrying staff in hand, and their belongings such as bowls hanging from either ends of a pole carried on the shoulder, are one of the most favourite themes" depicted on the plaques.31 Here are two unmistakable motifs ultimately derived from the Harappan, of the ‘emaciated or skeletal body' (Sign 48) and the ‘yoke-bearer' (Signs 12-15). It is significant that the two Indus ideograms are found paired in the Indus Texts (Fig.12).

An extraordinary plaque from Paharpur combines both motifs in one composite figure (Pl. VI),32 reminiscent of the technique of composite signs in the Indus Script. The plaque depicts a naked ascetic reduced to a skeleton with a bent back and exaggeratedly prominent ribcage and backbone and folded legs. He is carrying a ladle in his right hand (cf. the Kalibangan seal described above). He is also carrying a yoke on his shoulders to which are tied at either end a pair of vessels with ropes.

It is possible to interpret the Paharpur plaques depicting sepearately the ‘yoke-bearer' and the ‘skeletal ascetic' in terms of the BrahmĀnical legends of Bharadvaja (lit., ‘the bearer of victuals') and Dadhica (famous for his gift of his own ribs and bones) respectively. This is indeed more likely as the Harappan symbols from which they are ultimately derived would have been long forgotten when these plaques were made. However the extraordinary combination of the two motifs in one composite figure strengthens the hypothesis that they are the survivals of the two related motifs depicted in the Indus ideograms (Nos.15 and 48).

Section III: Identification of the Harappan Skeletal Deity with Dr. *Muruku

3.1 We have so far looked at the pictorial depictions of the ‘Harappan Skeletal Deity' and the ‘Emaciated Ascetics' of the later mythology and art tradition to learn what we can about the external attributes of the deity. An attempt will be made in this section to discover the probable original name and nature of the deity by searching through the Dravidian (Dr.) etyma with the nearest meanings corresponding to the pictorial elements.33

3.2 As seen earlier, the two defining characteristics of the pictorial depictions of the Harappan deity are (a) a skeletal body, and (b) bent and contracted posture. The Dr. etyma with the nearest meanings are as follows.34

(a) ‘To be shrivelled' (DEDR 4972):

Ma. muratuka: to shrivel; muraluka: id., decay.

Ka. muratu, murutu, muruntu: shrink, shrivel.

Tu. muruntu: shrunk, shrivelled.

Nk. mural: to wither.

Kur. murdna: to be dried to excess.

(b) ‘To be contracted' (DEDR 4977):

Ta. muri: to bend; murivu: contracting, fold; muri (nimir): (to stretch by) winding limbs.

Ka. murige: bending, twisting; muruhu: a bend, curve, a crooked object;

Ka. muratu, murutu, muruntu: to be bent or drawn together, state of being contracted.(DEDR 4972).

Tu. muri: curve, twist; murige: twist.

Pa. murg: to be bent; murgal: hunchback.

Ga. murg: to bend; murgen: bent; murug: to bend down.

Go. moorga: humpbacked.

(cf: Pkt. muria: twisted; old Mar. mured: to twist.)

We may infer from the linguistic data summarised in (a) and (b) that PDr. * mur/mur-V is the primitive root from which words with the meanings ‘shrivelled' and ‘contracted' have been derived.

3.3 We may now proceed to apply the technique of rebus to try and discover the Dr. homonyms with the intended meanings.

(c) ‘Strong, fierce, wild, fighting' (DEDR 4971) :

Ta. muratu: ill-temper, wildness, rudeness; muran: fight, battle, fierceness, strength.

Ma. muran: fight, strength.

Ko. mort: violence (of action); mordn: violent man.

Tu. murle: quarrelsome man.

Te. moratu: rude man.

(d) ‘To destroy, kill' (DEDR.4975) :

Ta. murukku: to destroy, kill; murunku : to be destroyed.

Ma. muruka: to cut.

Kol., Nk. murk: to split, break.

Kui. mrupka: to kill, murder.

Kur. murukna: to mangle, mutilate.

Malt. murke: to cut into bits.

(e) ‘Ancient' (DEDR. 4969) :

Ta. murancu: to be old, ancient; muri: antiquity.

Kol., Pa. murtal: old woman.

Nk: murtal : old woman.

Go. mur-: to mature.

The two sets of etyma in (c) and (d) taken together indicate that the original name of the deity was something like * mur/mur-V and that his essential traits were those of a fierce god, destroyer or hunter.

3.4 The legends and myths surrounding the deity have become inextricably mixed up and both sets of etyma in groups (a) to (d) apply to him. In short, the deity was both ‘a departed soul or demon' as indicated by his skeletal body and contracted posture, and also ‘a fierce killer or hunter' as indicated by the Dr. etyma. Furthermore, the linguistic data in (e) can be interpreted to mean that the deity was considered to be ‘ancient' even in Harappan times.

3.5 In the concluding part of the Paper, we shall compare the traits of the Harappan Skeletal Deity as revealed by the pictorial depictions and linguistic data summarised above, with those of muruku (Murukan), the primitive god of the Tamils as recorded in the earliest layers of the Cankam poetry.35

3.6 The most striking aspect of muruku is that he had no form; he was a disembodied spirit or demon who manifested himself only by possessing his priest or a young maiden. When muruku possessed him, the priest (velan) went into a trance and performed the shamanic dance in a frenzy (veri atal). When muruku possessed the maiden (anankutal), her mother called in the priest (velan) to perform the veri dance to pacify the spirit and restore the girl to her senses.36

3.7 The second prominent trait of muruku was of a ‘wrathful killer' indicating his prowess as a war-god and hunter.37

3.8 The only physical traits which may be attributed to the primitive muruku are his red colour (cey) associated with blood and bloody sacrifices, and his spear (vel) associated with killing enemies and hunting animals. As muruku had no material body, these two physical traits are shown to belong to his priest, velan the ‘spear-bearer' who wore red clothes and offered red flowers in ritual worship involving the sacrifice of goats and fowls. There were no temples in the earliest times, and the worship was carried out in the open field (kalam) before a wooden altar.

3.9 Another very ancient aspect of the worship of Murukan, not alluded to in the Cankam poems, but strongly supported by Tamil tradition, is the ritual carrying of offerings on the kavati (yoke with the offerings tied to the ends by ropes). The Paharpur plaques noticed above may also be compared with the Tamil legends of muruku (the demon) and Itumpan, his kavati-bearing worshipper.38

3.10 Much of the later Tamil literature, and virtually all the Tamil inscriptions and iconographic motifs have been heavily influenced by the Sanskritic traditions of Skanda-Karttikeya-Kumara and have very little in common with the primitive muruku except the name Murukan.39 Even the meaning of his name has undergone a radical transformation from muruku ‘the demon or destroyer' to Murukan ‘the beautiful one', consistent with the later notion that gods must be ‘beautiful' and demons ‘ugly'. As P.L. Samy points out in his incisive study of Murukan in the Cankam works, there is no support for the later meaning in the earliest poems. He derives muruku (Murukan) and murukku ‘to destroy' from Dr. muru-, and endorses the identification of Murukan with muradeva (a class of demons) mentioned in the Rgveda, as proposed by Karmarkar.40

3.1 1 The muruku of the early Tamil society before the Age of Sanskritization was a primitive tribal god conceived as a ‘demon' who possessed people and as a ‘wrathful killer or hunter'. This characterisation of the earliest Tamil muruku is in complete accord with his descent from the Harappan Skeletal Deity with similar traits revealed through pictorial depictions, early myths and Dravidian linguistic data.


Notes and References

1. See, for example, the pictorial representations of deities and sacrificial scenes on seals, sealings and other inscribed objects. I. Mahadevan 1977 (=ISTCT), App. II: Field symbols 47-81; Pl.IV-VI: Figs. 80-116.

2. ISTCT, Sign List: Nos. 1-48 (p. 32).

3. Gardiner, Sign List: A.40 (determinative) and C.1-7 & 9-1 1: (ideograms).

4. It is interesting that Asko Parpola and his Finnish colleagues started off in 1970 with virtually the same assumptions and identified Sign 47 as ‘god' and Sign 48 as ‘Mother goddess'. Asko Parpola et al 1970: pp. 25-26. However Parpola has since changed his mind and presently identifies Sign 47 as an ‘enraged cobra with expanded hood' and Sign 48 as its ‘female gender'. Asko Parpola (1994: p.59).

5. ISTCT, Concordance, pp. 195-96. The pairing of the signs proves that they are not mere graphic variants. Either sign can function as a substantive or as an attribute. However when they occur as substantives, Sign 48 is followed by the ‘JAR' suffix (Sign 342) indicating masculine gender, and Sign 47 is followed by the ‘ARROW' suffix (Sign 211) indicating non-masculine gender. I. Mahadevan 1998.

6. Vats, vol. II (Plates): Figs. 303-613; 658-692. The votive tablets and sealings are interpreted as ‘sacrificial' by the Soviet scholars (transl. Zide & Zvelebil, pp.105-07); and as serving ‘a religious purpose' by Asko Parpola (1994: p.107).

7. ISTCT, Tables I & IV.

8. ISTCT, Concordance, pp. 197-200.

9. Joshi & Parpola (=CISI, Vol.I), No. K -102-03.

10. Sinha & Roy, p.121, Pl. XXX, No.24. A grey-coloured round terracotta seal with three Indus signs ( 47-342-176) to be read in the clock-wise direction starting from the 6 o'clock position (in the impression). This little-known seal was first identified as bearing a legend in the Indus Script by Chakraborty (p.88 & Pl. 3A). The excavators assign the seal to Period III (ca. 200 BC - 200 AD.). However it is difficult to believe that this seal (-bearing a text so similar to the Harappan that, had it been found at Harappa, it would not have attracted special attention-) can be so late. As the excavators point out, the site is a highly disturbed one, and PGW and NBP ware occur together "as the ware was re-deposited from the lower levels in the course of making the plinth of the Garh higher and erection of mud rampart" (Sinha & Roy, pp 7-8). Most probably the present seal came from the lowest level reached at this site (ca. 1100 BC).

11. Lal, Pl. XXXI B-1(Megalithic) symbol No.47. The symbol also occurs in Pl. IIIA-1,3 and Pl. XXX B-1. See especially Lal's photographic comparison of the Indus sign and the Megalithic symbol (Pl. XXXI B). He remarks: "In the case of Sanur (rare examples elsewhere also) three symbols occur in such close proximity to one another as to give the impression of a record. It may however be added that the three symbols interchange their positions on different pots producing all possible combinations" (Ibid. p.23). The graffiti-bearing Megalithic pottery found in Tamilnadu is assigned to ca. the second half of the First Millennium BC.

12. Gardiner, Sign List: F. 38-40.

13. CISI, Vol.I, No. K-104

14. CISI, Vol.I, No. K-48.

15. The details are clearly visible in the highly enlarged photograph of the seal published in Swami Oamanda Saraswati 1975, Pl. 275.

16. For illustrated Lists of variants of Sign 48, see ISTCT, p. 785, No.48; Asko Parpola 1994, p.71, No.87.

17. It has been suggested that Ta. pey is from Pkt. peya < Skt. preta. Filliozat 1982: p.10. Notwithstanding the weighty authority of DEDR (Entry 4438), I agree with Filliozat.

18. Ardeleanu -Jansen, pp. 139-57, Figs. 16-35.

19. The Vedic myths relating to Dadhyanca and Dadhikravan are summarised in Macdonell, pp.141-42 and 148-49. For references to Dadhica (Dadhici) in the Mahābhārata, see Sorensen, p.225.

20. Meriggi, pp. 198-241. Mishra, pp. 78-81.

21. Knorozov et al, Index of Signs, No.48 (pp. 84, 100).

22. Ellora, Cave 14.

23. Sorensen, p.147.

24. Gardiner, Sign List: A-6. Wallis Budge, Sign List: 1.101-104.

25. Debala Mitra, Pl.5.

26. Karaikkal Ammaiyar, Arputa-t-tiruvantati 101. Filliozat, pp. 10-11.

27. Karaikkal Ammaiyar, Tiruvalankattu mutta tiruppatikam.

28. Bronze of Karaikkal Ammaiyar (Chola Period) at Bhava Aushadhisvara temple in Tirutturaippundi (Thanjavur Dt.).

29. S.L. Shali, pp. 133-34.

30. Emaciated Ascetic, stamped terracotta tile, 4-5 cent. AD. Harwan. (Prince of Wales Museum, Bombay). A similar frieze from this site in the Ashmolean Museum, Oxford, UK, (Mus. Acc.No.1980-65) depicts additionally the ribs. See also Debala Mitra, Pl. 85.

31. K.N. Dikshit, p.66; Pl. XXVI (b): (yoke-bearer); Pl. XLVIII(e) : (Emaciated Ascetic).

32. Ibid Pl. XXVI (a). The photograph published in the book is not clear. Pl. VI illustrating the present Paper is from a much better photograph (ASI. 16/64) in the Photo Archives of the Archaeological Survey of India, New Delhi.

33. The basic premise is that the Indus Texts are in a Dravidian language. The arguments in favour of the Dravidian character of the Indus Valley Civilization are presented in Parpola 1994, pp. 160-75.

34. The Dravidian linguistic data is taken from Burrow & Emeneau, A Dravidian Etymological Dictionary, 1984 (=DEDR). Names of Dr. languages are abbreviated as in DEDR.

35. The earliest layer of the Tamil Cankam poetry comprises the Ettuttokai (Eight Collections) and Pattuppattu (Ten Idylls) excluding Tirumurukarruppatai and Paripatal which are considered to be relatively later works. For analysis of the earliest references to Murukan in the Cankam literature, see P.L. Samy. For an overall view of Tamil Traditions on Subrahmanya-Murukan, see Kamil Zvelebil 1981 & 1991.

36. Akam. 22, 98, 138, 139 etc.

37. Akam. 59, 158, 266; Puram. 14, 16; Narrinai. 225 etc.

38. See Asko Parpola 1981 & 1997 for the connection between the Indus sign ‘yoke- carrier', kavati traditions in North India as reflected in Indo-Aryan languages and kavati worship in Tamilnadu.

39. The earliest epigraphic reference to Murukan in Tamilnadu is found in the Tiruttani (Velancheri) Plates of Pallava Aparajitavarman (ca. 900AD); R. Nagaswamy. Sculptures of Murukan begin to appear only from the Pallava-Early Pandya Period (from ca. 7-8 cent. AD). For a comprehensive treatment of the iconography of Murukan in Tamilnadu, see L'Hernault.

40. P.L. Samy, pp. 9-16, 96. A.P. Karmarkar, p.128. it is significant that the name mura in the RV is derived by Sayana from the root with the meaning marana ‘killing'.


Ardeleanu-Jansen, Alexandra 1983. Stone sculptures from Mohenjo-Daro. Interim Reports, Vol.I. ISMEO- Aachen University Mission. Aachen.pp.139-157.

Budge, Wallis. 1966. Egyptian Language. London.

Burrow T. & M.B. Emeneau 1984. A Dravidian Etymological Dictionary. Second Edition (= DEDR). Oxford.

Chakraborty B.B. 1981. Indus Script: A Further Light. Calcutta.

Debala Mitra 1971. Buddhist Monuments. Sahitya Samsad. Calcutta.

Dikshit K.N. 1938. Excavations at Paharpur, Bengal. Memoirs of the Archaeological Survey of India, No.55. Delhi.

Filliozat J. 1982. (Introduction in) Karaikkalammaiyar. Nouvelle Edition. Institut Francais D'Indologie. Pondichery.

Gardiner A. 1973. Egyptian Grammar. Oxford University Press. London.

Joshi J.P. & Asko Parpola 1987. Corpus of Indus Seals and Inscriptions (=CISI). Vol. I: Collections in India. Helsinki.

Karmarkar A.P. 1950. The Vratya Dravidian systems. Lonavala.

Knorozov Y.V. et al 1981. Proto-Indica 1979. Moscow.

Lal B.B. 1960. From Megalithic to the Harappa: tracing back the graffiti on pottery. Ancient India, 16, pp. 4-24.

L'Hernault F. 1978. L'Iconographie de Subrahmanya au Tamilnad, Institut Francais d' Indologie. Pondichery.

Macdonell A.A. 1981. Vedic Mythology. Motilal Banarsidass, Delhi.

Mahadevan I. 1977. The Indus Script: Texts, Concordance and Tables (=ISTCT). Archaeological Survey of India. New Delhi.

__________ 1998. Phonetic value of the ‘ARROW' sign in the Indus Script. Journal of the Institute of Asian Studies. Vol. XV, No.2, pp.69-74.

Marshall, John 1931. Mohenjodaro and the Indus Civilization. 3 Vols. London.

Meriggi P. 1934. Zur Indus-Schrift. ZDMG, 87 (NF 12). pp. 198-241.

Mishra, Madhusudan 1998. From Indus to Sanskrit, Pt.III. Delhi.

Nagaswamy R. 1979. Thiruttani and Velanjeri Copper Plates. State Dept. Of Archaeology, Tamilnadu. Madras.

Omananda Saraswati 1975. Ancient Seals of Haryana (in Hindi). Rohtak.

Parpola A. 1981. On the Harappan ‘Yoke-Carrier' pictogram and the Kavadi worship. Procds. of the Fifth International Conference-Seminar of Tamil Studies. Madurai. Vol. I: pp. 2.73-89.

__________ 1994. Deciphering the Indus Script. Cambridge University Press.

__________ 1997. Dravidian and the Indus Script: On the interpretation of some pivotal signs. Studia Orientalia 82, pp. 178-87.

Parpola A. et al 1970. Further Progress in the Indus Script Decipherment. The Scandinavian Institute of Asian Studies, Copenhagen.

Samy P.L. 1990. Cankanulkalil Murukan (Murukan in the Cankam Works). Madras.

Shali S.L. 1993. Kashmir: History and Archaeology through the Ages. New Delhi.

Sinha B.P. & Sita Ram Roy 1969. Vaisali Excavations (1958-1962). Directorate of Archaeology and Museums. Patna.

Sorensen S. 1963. Index to the Names in the Mahābhārata. Motilal Banarsidass. Delhi.

Zide A. & K.V. Zvelebil 1976. The Soviet Decipherment of the Indus Valley Script: Translation and Critique. Mouton. The Hague.

Zvelebil K.V. 1981. Tiru Murugan. International Institute of Tamil Studies, Madras.

__________ 1991. Tamil Traditions of Subrahmanya-Murugan. Institute of Asian Studies, Madras.

This article first appeared in the March 1999 issue of The Journal of the Institute of Asian Studies.

Iravatham Mahadevan, I.A.S. (Ret'd) joined the Indian Administrative Service in 1954 and took voluntary retirement in 1980 to devote himself exclusively to academic pursuits. Among his contributions to Indian epigraphy, especially in the fields of the Indus and Brahmi scripts, his book The Thdus Script: Texts, Concordance and Tables (1977) is recognised as a source-book for research in the Indus script. His most recent book is Early Tamil Paleography .

"Deciphering the Indus Script" (book review) by Richard Solomon
"Indus Graffiti as Rock Art and their Astronomical Implications"

Index of research articles on Skanda-Murukan