1 Tantra is esoteric since its adepts, tāntrikas, are supposed to have been initiated. The sectarian initiation, dīkṣā, being held as a first step towards liberation. All Tantric adepts or believers, however, are not initiated, But with or without initiation Tantric teaching is always described as secret. On the role of secrecy, see studies by Hugh Urban, notably The Economics of Ecstasy. Tantra, Secrecy and Power in Colonial Bengal (New York: Oxford University Press, 2001).
2 In some cases, as we shall see (for mantroddhāra, for instance – see chapter 1, pp. 12ff.), mantras are to be written, or to be visualised mentally in their written form, but these are exceptions to the general rule of orality.
3 A most common form of those mantras – a form current in Tantric practice too – is for instance OṂ HAUṂ śivāya namaḥ.
For an overview of mantras in Hinduism, see chapter 22 in G. Flood, ed., The Blackwell Companion to Hinduism (Oxford: Blackwell, 2003). On mantras generally, the only comprehensive study remains to this day H. Alper, ed. Understanding Mantras (Albany: SUNY Press, 1989).
4 The word mantra is of the masculine grammatical gender.
5 On these exclamations – or inflections, as they are technically called – see the entry jāti of the Tāntrikābhidhānakośa (TAK), vol. 2.
6 On these, see A. Padoux, Vāc. The Concept of the Word in Selected Hindu Tantras (Albany, SUNY Press, 1990, repr. Delhi, Sri Satguru Publications, 1992); this, however, being a translation of a 1964/1975 French thesis, reflects the state of knowledge of the subject at a time when many tantras now being studied were unknown or not available.
7 These alphabets are also forms of Bhairava or of the goddess: they are alphabetdeities.
8 ‘The mātr̥kā, the unknown universal Mother who gives birth to the cosmos through the [Sanskrit] phonemes from A to KṢA’, to quote from Kṣemarāja’s commentary on Śiva Sutrā 1.4. See here, chapter 4, on nyāsa: the ritual placing of the alphabet is called mātr̥kānyāsa. The conception of the goddess Mātr̥kā and of her role is not the same in all tantras. This subject is being studied by Judit Törzsök: paper read in Pondicherry in 2009.
9 For instance, Ajtāgama 1.13–14, the tantrāvatāra chapter, on the ‘descent’ of the tantra, not a chapter on mantras.
10 Not so surprising, however, if one refers to the Vedic stobhas, such as ī hā bu hvā.
11 pārameśvare’pi avyaktadhvaner mukhyatayaiva prāyaśo mantratvam nirūpitam. This work has been edited with an Italian translation by R. Gnoli (Roma, IsMEO, 1985). There also exists a (less scientific, and difficult to use) English translation by Jaidev Singh (Delhi, Motilal Banarsidass, 1988) – a new edition is being prepared by Bettina Bäumer).
12 The utterance of the mantra, that is. On uccāra, see A. Padoux, Vāc The Concept of the Word in Selected Hindu Tantras, chapter 7; or the entry uccāra of the TĀK, vol. 1. See too, here, chapter 3 on japa. In Tantric usage, the uccāra of a mantra is often a yogic practice where the mantra moves upward together with the kuṇḋalinī – such is the case of the japa of the śrīvidyā described here, pp. 40–41. On this subject, see A. Padoux, ‘Corps et mantra. la présence des mantras dans le corps’, in O. Botto, et al., eds., Du corps humain au carrefour de plusieurs savoirs en Inde (Bucarest-Paris, de Boccard, 2004, Studia Asiatica vols. IV–V), pp. 563–578.
13 This was already stated in the Laws of Manu (2.85) and remains valid to this day. See chapter 3, on japa, p. 25.
14 It may be interesting to quote here the Tamil poet Tirumular’s (c. seventh century) definition of mantra as ‘perfect concentration of the mind on anything’.
15 In the dualistic Śaivasiddhānta, mantras are often called aṇu, that is, atom or rather individual soul. Mantras appear also as discrete, different, entities in the system of pramātr̥.
16 Such is the case, for instance, of the japa of the śrīvidyā which takes place at the end of the cult of Tripurasundarī described in chapter 3 of the Yoginīhr̥daya (YH) where the śrīvidyā, which is the mantric aspect of the goddess, ascends along the suṣumnā in the body of the worshipper, or of the ‘subtle’ meditation of the Netramantra described in chapter 7 (pp. 15–51), both briefly described here, chapter 3.
17 mantrā varṇabhattārakā laukikapārameśvarādirūpā mananatrāṇarūpā vikalpasaṃ vinmayyāḥ. (PTV, p. 243).
18 Nor were they the same in the śaivāgamas and in the earlier Pāñcarātra saṃhitās, and still less in the later ones.
19 These elements were taken over from the indexes of Vedic hymns (anukramaṇikā). They are completely arbitrary.
20 It is fair to say, however, that mudrās, as parts of a ritual or used otherwise, are sometimes held to be by themselves effective. This is the case in many rituals where, for instance, a deity is invoked and placed on a ritual support with a mudrā. Such is also the case of the mudrās described in chapter 1 of the Yoginīhr̥daya (1.56–71). These mudrās are also aspects of the power of the goddess, and deities worshipped during the pūjā. On this, see chapter 4.
21 Though japa, especially when made using the rosary, includes rites – see here chapter 5.
22 This is done by the astra (weapon or arrow) mantra PHAṬ or the kavaca (cuirass) mantra.
23 In the prāṇapratiṣṭhā ritual.
24 They may therefore be described as performative utterances or as illocutionary forces, in the sense of J.L. Austin.
25 Then come, in decreasing importance and value, the vijñānākalas, the pralayākalas and finally the sakalas. On this system, see H. Brunner’s ‘Mantras et mantras dans les tantras śivaïtes’, in R. Torella, ed., Le parole e i marmi, Studi in onore di Raniero Gnoli (Roma, IsIAO, 2001), pp. 185ff., or Vāc, p. 104 n.
26 Which was the case of vāc in the Veda: ‘the Word that speaks and is the harmonious ruler of the gods’ (R̥V 8.100,10).
27 My position differs in this from Hélène Brunner’s in her 2001 study referred to in note 25, where she distinguishes clearly betwen ‘Mantras’ as powers and ‘mantras’ as sound units, this distinction corresponding to two different uses of mantras. The distinction of uses, I believe, does not necessarily imply a distinction of nature.
28 The mūlamantra of a deity is the deity itself. The mūrtimantra evokes the deity’s form; the aṅgamantras of a deity symbolise and evoke the aṅgas, the ‘limbs’, that is, the constitutive parts or qualities (guṇa) of that deity. The ‘weapons’ (āyudha) of a deity have also mantras used to evoke them or to place them on the deity’s image.
29 See the uccāra of OṂ, SAUḤ or of the saṃhārabīja KHPREṂ/HSHPHREṂ described in Vāc, pp. 401–426.
30 dhyāna in Tantric context is the word for the mental visualisation of a deity – in addition, of course, to its usual meaning of meditation.
31 On this subject, see chapter 2.
32 Two such well-known manuals of iconography are Mahīdhara’s Mantramahodadhi (sixteenth century) and the somewhat later Mantramahārṇava. The former has been translated with introduction and notes by G. Bühnemann, The Iconography of Hindu Tantric Deities, vol. 1 (Groningen: Egbert Forsten, 2000). The second volume deals with the pantheons of the Prapañcasāra and of the Śāradātilaka (ibid., 2001).
33 This ritual is described in several texts, notably in chapter 15 of the Kulārṇavatantra (KT), on which see G. Bühnemann’s study in T. Goudriaan, ed., Ritual and Speculation in Early Tantrism (Albany: SUNY Press, 1992), pp. 61–106.
34 The rituals thus performed are therefore classified among the kāmya, ‘desiderative’ rites.
35 On these see T. Goudriaan, Māyā Divine and Human (Delhi: Motilal Banarsidass, 1978), pp. 251–412.
36 This subject is approached more fully in chapter 7, on mantras, of Vāc, quoted in note 12.
37 Or should we say ‘mantra-deity’? Or ‘mantra qua deity’?
38 tasya yā sattā aśeṣaviśvābhedamayapūrṇāhaṃvimarśatmā sphurattā sā mantrāṇāṃ rahasyam upaniṣat (ŚSV 2.3, comm., p. 50 of the KSTS edition, Srinagar, 1911).
39 In this perspective, mantras, made of phonemes ‘extracted’ from the Sanskrit alphabet, are also (or at least one of them can be) at the source of the alphabet.
40 tadākramya balaṃmantrāḥ sarvajñabalaśalināḥ/
pravartante ‘dhikārāya karaṇām īva dehinām //26//
tatraiva saṃpralīyante śāntarūpā nirañjanāḥ /
sahārādhakacittena tenaite śivadharminaḥ // 27//
41 This, the Saiddhantika authors would see differently since for them the mantras are powers that act by themselves. Their position is explained by H. Brunner in the article quoted in note 25.
42 mahāhradānusaṃdhānān mantravīryānubhavaḥ /.
43 See infra, p. 58 for nyāsa or p. 46 for japa.
44 New York: Peter Lang, 1990/93; Indian edition, Delhi, Motilal Banarsidass, 1996.
45 Vedic recitation, too, was governed by the rule of the absolute respect of the phonetic content of the recited mantra. The Vedic reciter concentrates on the sounds he emits, not on the meaning of the words he utters: this is still the case today. The Tantrika is heir to the Vedic r̥ṣi.
46 Such views are briefly reviewed in A. Padoux, L’énergie de la parole, pp. 178–190 (Paris, Le Soleil Noir, 1980, reprint Fata Morgana, 1994).
47 On mantras in general I would refer the reader to Understanding Mantras, quoted in note 3 and, perhaps, more specially, to the article I wrote as a (far from definitive) conclusion to that book: ‘Mantras, what are they?’ (pp. 295–318).
1 This is a new version of a paper ‘La sélection des mantras’ published in the BEFEO, LXV, 1978, pp. 65–85.
2 As is often said: mantrāḥvarṇātmakāḥsarve varnāḥ sarve śivātmakāḥ, ‘all mantras are made of phonemes which are all made up of Śiva’.
3 vidyās are feminine mantras (or female mantradevatās), or mantras of female deities. On the nature of mantras, see the Introduction.
4 See Padoux, Vāc, chapter 1 on vāc, and chapter 7 on mantras.
5 On the term uddhāra, see the entry on this word in vol. 1 of the Tāntrikābhidhānakośa (henceforth referred to as TAK), p. 231.
6 In chapter VIII of Teun Goudriaan and Sanjukta Gupta, Hindu Tantric and Śākta Literature (Leiden, Brill, 1981).
7 A well-known one is Dakṣiṇāmūrti’s Uddhārakośa, edited by Raghu Vira and Shodo Taki, Delhi, 1978. Or the Tantrābhidhāna with Vīja-nighantu, edited by Tāranātha Vidyāratna (‘Tantrik Texts’ vol.1, Calcutta 1913).
8 The reason for this identification, as well as that for M and time, is not clear for me.
9 caturdaśayutaṃ bhadre tithīśāntasamanvitam//
tritīyaṃ brahma suśroni hr̥dayaṃbhairatmanaḥ:
10 In the varṇasamāmnāya (notably in the case of the phonematic emanation, the varṇ aparāmarśa as expounded in the Tantrāloka (TĀ)), the additional phonetic signs bindu and visarga are included among the vowels which are thus considered as being sixteen in number. (I follow here the traditional practice of considering the diphtongs as vowels. See Padoux, Vāc, pp. 223ff.).
11 tritīyaṃbrahma sakāraḥ, says Jayaratha ad TĀ 30–.167 – this because of Bhāgavadgīta 17.23: oṃ tat sad iti nirdeśo brahmaṇas trividhaḥ smr̥tah, where sad (therefore, here, SA or S) is quoted as the third of the three phonemes (oṃ tat sat) that are the brahman. The esoteric play on words goes here quite far, but the more obscure, the better! (The use of words in more than one sense (śleṣa) is a current Indian traditional interpretative practice – on the subject, see for instance Sanderson’s remarkable study, ‘A commentary on the opening verses of the Tantrasāra of Abhinavagupta’, in S. Das and E. Fürlinger, eds. Sāmarasya, Studies in Indian Arts (New Delhi, D.K. Printworld, 2005) pp. 89–148.
12 See the articles ācārya, guru and deśika in vols. 1 and 2 of TAK.
13 On this process, see Hélène Brunner’s article, ‘Le sādhaka, personnage oublié du śivaïsme du sud’ (JAs, 1975, pp. 411–443, and Gudrun Buhnemann, ‘On Puraścaraṇa: Kulārṇavatantra, Chapter 15’ in T. Goudriaan ed., Ritual and Speculation in Early Tantrism (Albany, SUNY Press, 1992), pp. 61–106.
14 Admittedly, alphabet is not the best term to name the ensemble of the 49/50 Sanskrit phonemes, varṇas. I use it however for simplicity’s sake since it is familiar to us. As is well known, the Sanskrit alphabet includes 14 vowels (svara) and two additional phonetic units, 4 semivowels (antaḥstha), 25 consonants (vyañjana), and 4 sibilants, a total of 49 letters to which, in Tantric traditions, is added the consonant group KṢA.
15 The whole alphabet is also sometimes considered as a mantra. There are, too, alphabet-goddesses.
16 Note to his edition of the Mahānirvānatantra, III, 11–12 (p. 33), a passage where the mantroddhāra is very much simplified, but Avalon’s formula is nevertheless generally valid. Arthur Avalon was no expert in Sanskrit, but the pandits he worked with were. His contribution to the Tantric field was important – and remains useful. The Mahānirvāṇatantra is a late (18th c.) fabrication. It includes elements showing British influences, but it is still worth quoting on some specific points.
17 We shall see later on the technical meaning of this term.
18 varṇasaṃghaṭtanāśarīrāṇāṃ mantrāṇāṃ saiva bhagavatī vyākhyātarūpā vidyāśarīrasattā rahasyaṃ iti pradarśitam/pratyāgamaṃ ca mātr̥kāmālinīprastārapūrvakaṃ mantroddhārakathanasya ayam eva āśayaḥ.
On mātr̥kā and mālinī, see Padoux, Vāc, pp. 312–327.
19 A Śaiva text; one of the early (perhaps eighth century) Bhairavatantras. It was commented upon, according to the non-dualist Trika system, by Abhinavagupta’s disciple Kṣemarāja (the Uddyota). Text and commentary are published in the ‘Kashmir Series of Texts and Studies’ (Bombay, 1921–1933).
20 mātr̥kāṃ paśunām ajñātāṃ viśvamātaraṃ sarvamantratantrajananīṃ ādikṣābtām … (SvT-Uddyota, vol. 1, p. 26). See Śivasūtravimarśinī 1.4 (p. 16): tasya ādiksāntarūpā ajñātā mātā mātr̥kā viśvajananī: ‘its unknown mother, mātr̥kā, begetter of the universe, formed of [the phonemes] from A to KṢA’.
21 prastaret, says SvT 1.31b, verbal form explained by Kṣemarāja (ibid., p. 277, l.1) as an ārṣa form for prastr̥ṇīyāt.
22 ekacittaḥ is explained in the Uddyota (vol. 1, p. 26) as mantrodayanibhālanapravaṇaḥ, intent upon the perception of the production (creation or coming up) of the mantra.
23 The Mahāmātr̥kā or Mahāmātaraḥ are in fact usually seven: the so-called Saptamātr̥kā – and this is the number mentioned in this passage of the SvT and in Kṣemarāja’s Uddyota, in spite of the fact that the text actually enumerates not seven but eight goddesses.
24 The mūrtimantra is the mantra which, in the ritual, serves to invoke the subtle bodily aspect (mūrti) of the deity being worshipped – or which is, or represents, on a subtle plane, the body of the deity.
25 ante’syā uddharen mantrān yathākramaniyogataḥ// 37//
Comm.: asyā – mātr̥kāpūjāyāḥ, yathakramam – āsanamantramūrtimantrādikramena yo niyogaḥ – tattanmantramūrtimantrādikramena yo niyogaḥ – tattanmantrāvācya devatānusaṃdhānaparatvaṃ tena.
26 See for instance Rauravāgama, kriyāpada, 1.1–21.
On the notion of aṇga, I refer to Hélène Brunner’s study, ‘Les membres de Śiva’, Asiatische Studien/Etudes Asiatiques, XL-2, 1986, pp. 90–132.
27 This is composed usually of milk, curds, ghee, cow urine and dung. Another group of five offering substances is that called pañcāmr̥ta, which, in ordinary rituals, consists of five ‘innocuous’ substances such as molasses, honey, ghee, curds and milk, but, in Kaula rites, is made up of human fluids and excreta mixed with alcoholic liquor (vāmāmr̥ta). On these, see A. Sanderson, op. cit., note 63, p. 113.
28 The officiating person is, if not an ācārya, at least an initiated putraka or sādhaka. Note that here as in the case of the NT, the adept is to act ‘with devotion’: the manipulation of the sacred does not exclude faith, devotion, nor does it exclude its correlate, divine grace.
29 In the Pāñcarātra tradition (as in other ones), each of the fifty phonemes – as a phoneme and as a deity – has a name or names. A may thus be called aprameya, prathama, viṣṇu, etc. Śl. 32 to 58 of this chapter enumerates these names. Such lists of names are given as appendices in the Sanskrit editions of the Jayākhyasaṃhitā or of the LT (see also app. XIII of the English translation of that tantra). See too Dakṣiṇamūrti’s Uddhārakośa, or chapter 347 of the Agnipurāṇa (ekākṣarābhidhānam.).
30 The LT, chapter 25, śl. 30–37 (p. 86, p. 136–137 of the English translation) expresses itself in a similar fashion: ‘The phonemes made powerful by the śakti of Lakṣmī are the source of all mantras. They must be intensely worshipped, praised and meditated upon using their names …, they participate mutually in each other when used in mantras. There is nothing in this movable and immovable world that is not produced by them. Although the forms of mantras are supreme, divine and eternal, they are yet conceived of as produced by the phonemes … Just as one imagines divisions in space which is one, so, for convenience’s sake, one imagines that there are divisions between phonemes [though they are indivisible sound-units] … Having thus clearly realized in his mind this presence of the phonemes in the mantras, and having worshipped them on the ground, in the lotus [the diagram] and in the body of the goddess, he should extract them (mantrān samuddharet)’. The LT, chapters 23–24, describes in details how to display the mātr̥kā, the mother of the mantras (mantramātr̥kā), the divine power creating and pervading the cosmos, and how to worship her before extracting the mantras. The tantra adds that once the lotus-shaped cakra has been drawn, the adept is to visualize the mantramātr̥kā, who is the Self of Viṣṇu as seated on the pericarp of the lotus, her body being made of the fifty phonemes. He is to worship her before extracting the mantras. The Sanatkumārasaṃhitā (Indrarātra, chapter 2), gives a similar description.
31 A good summary of this passage is given by Marion Rastelli, Philosophisch-theologische Grundanschauungen der Jayākhyāsaṃhitā (Vienne, 1999), pp. 126–128.
32 There are fifty-one letters when the vedic l̥ is added after KṢA.
33 LT 24, 1–34, pp. 80–82; pp. 127–130 of the English translation.
34 In the LT, chapters 23 and 24, there is a very complete and interesting description of how to display the mātr̥kā, the mother of mantras (mantramātr̥kā, the divine energy pervading the cosmos she has begotten, and how to worship her before extracting the mantras. The tantra adds that once the diagram has been drawn, one must visualise the mantramātr̥kā who is the very Self of Viṣṇu as being on the pericarp of the lotus, her body adorned by and made of the fifty phonemes. She is to be worshipped by the master who must identify himself with her before extracting the mantras (LT, pp. 77–81). The Sanatkumārasaṃhitāa (Indrarātra, chapter 2: mantroddhāra, pp. 228ff. of the Adyar, 1969 edition) gives a largely similar description.
35 On mātr̥kā and śabdarāśi as on mātr̥kā and mālinī, see Padoux, Vāc, chapter 5, pp. 305–329.
36 On the prastāras and gahvaras, see Appendix I of The Ṣaṭsāhasra Saṃhitā, chapters 1–5 (Leiden, 1982), edited and translated by J.A. Schoterman: it is a very complete and interesting study of the prastāra and gahvara system, based on an as yet unedited text of the Kubjikāmata Kaula tradition, the Saṃvartārthaprekāśa of Mukundarāja, where a large number of such diagrams are reconstructed.
Appendix II of Schoterman’s study (pp. 210–221) is on Mālinī, both as an order of letters and as a goddess, and on Śabdarāśi, both as the Sanskrit alphabet and as the god Śabdarāśibhairava.
The fourth chapter of the Kubjikiāmatatantra – KMT (75–106) describes the extraction of mantras from the mālinīgahvara. mālinī being feminine, it is the energy aspect of phonemes that prevails here. The goddess Mālinī is said in the tantra to be the supreme matrix (parā yoni) of the Rudras and Yoginīs.
There is a Yonigahvaratantra where a gahvara is said to be the matrix (yoni) of all mantras. Gahvaras are also described in the Gorakṣasaṃhitā.
37 One finds frequently in India prescribed processes that are as theoretically or apparently constraining as they are in actual practice flexible and adaptable. This is met with continually in Āgamic ritual prescriptions. But it corresponds, in fact, to a very general attitude in Hinduism, where, very often, the general prescribed ruling is accompanied by so many circumstantial precisions or exceptions that it can eventually be adapted to any issue. In the same way, in purāṇic hinduism, to quote Madeleine Biardeau (‘Etudes de mythologie hindoue’ BEFEO, LVIII, pp. 63–65), ‘while the official aim is to be freed of rebirth, all functions so as to allow the world to go on existing eternally’. This seems to be a characteristic trait of the Hindu religious-social consciousness, one of the ways in which Hinduism integrates and resolves its contradictions.
38 Two texts that I happen to have at my disposal: others could as well have been quoted.
39 Pp. 1–26, vol. 4 of the KSTS edition of the SvT.
40 This is Ksemarāja’s gloss.
41 śl. 3 to 10 enumerate various deities or supernatural beings of whom the adept might be an amśa and the efffect this will have. Thus the adept who has brahmāmṣa will be devoted to the Veda, with rudrāṃśa he will be rudrabhakta; the adept with amśa of the Gandharvas will sing constantly, that of Kāma will love courtesans (gaṇikapriya), and so forth. In all these cases, explains Kṣemarāja, the deity will give the adept the fruit corresponding to its power after being meditated upon, its mantra being recited and being offered an oblation (svasiddheḥ phaladāḥ sarve svadhyānajapahomataḥ – śl. 11a).
42 Pp. 8–9, vol. 4; the process is not described in the Tantra but in the commentary.
43 There are many periphrases for designating human flesh used as ritual offering. Used here are the two rather unusual terms vaihāyasa and dhvaja, which, according to Kṣemarāja’s explanation, are used as meaning [the flesh] of a man who has been hanged (udbaddha), and of a man who has been executed by impalement (śūlāropita). However, as Professor Alexis Sanderson (whose help in this matter I gratefully acknowledge) suggests Kṣemarāja may have interchanged the meanings of these two terms which are sometimes to be met with, the two sorts of human flesh being paired for option, in Hindu and Buddhist tantras.
Common terms for sacrificial human flesh are mahāmāṃsa (great flesh), mahādhūpa (great incense), mahākṣmā, etc.
44 The same practice is described briefly (and more clearly) in Kṣemarāja’s commentary of NT 8. 59b–63, vol.1, pp. 208–209.
45 On this see for instance the ŚT, 2. 127–128 and its commentary (vol. 1, p. 102). This cakra is shown and explained in G. Bühnemann’s study of puraścaraṇa in T. Goudriaan, ed., Ritual and Speculation in early Tantrism. Albany: SUNY Press, 1992, p. 100.
46 This cakra is mentioned in the KT 15, 86. It is reproduced on p. 97 of G. Bühnemann’s study quoted above.
47 It is shown and explained in Bühnemann, op. cit., p. 101.
48 See Bühnemann, op. cit., p. 96.
On all these methods, see the commentary of ŚT, 2.129–131 (vol. 1, pp. 102–112) which refers and quotes several tantras: MVT, Mantramukāvali, and (frequently) Piṇgalāmata. The Tārabhaktisuddhārṇava (TBhS), a seventeenth century manual, in its section on aṃśa, quotes in its tenth chapter (pp. 359–366) many such texts.
Krsṇa-vaiṣṇava practices of this sort (very similar to the Śaiva ones) are described, in French, with reference mainly to the Rāmārcaṇcandrikā, in R.V. Joshi, Rituel de la dévotion kr̥ṣṇaïte (Pondichéry, 1959), pp. 20–22. The diagram with four squares is shown there. This diagram and those used in other aṃśa/ka practices (often called śuddhi in the sense of discharge) are to be found in most contemporary ritual manuals, in all Indian tongues. For example, in Hindi: Mantrasiddhi kā upāya, by Bhadraśīla Śarma (Allahabad, 1968).
49 The kūrmacakra is described in the ŚT, 2.133–137 and in pp. 114–117 (vol. 2) of the commentary, without, however, the way to use it being clearly explained. This is an important diagram used also in other ritual circumstances, in puraścaraṇa notably. It is also described in the short chapter 20 of the Pheṭkāriṇītantra, where it is explained how to use it to determine the aṃśa; the tantra also mentions the mantras whose validity has not been checked.
50 It is sometimes also called mantrārādhana. The term puraścaraṇa can also be applied to the worship of the mantra – sādhanaṃ mūlamantrasya puraścaraṇam ucyate, says a text quoted in the TBhS, p. 315 – since it is the preliminary act (puraskriyā) of the mastery of the mantra. On puraścaraṇa, see G. Bühnemann’s study ‘On Puraścaraṇa: Kulārṇavatantra, chapter 15’, in T. Goudriaan, ed., Ritual and Speculation in Early Tantrism.
51 See for instance the cult of two Yoginīs prescribed in YH 3 166.
52 The SvT, in chapter 8, śl. 3–10, enumerates the different powers the adept will possess if he has the aṃśa of the mantra of various deities: see above, note 36.
53 On these magical acts, see T. Goudriaan, Māyā Divine and Human (Delhi: Motilal Banarsidass, 1978), which is according to its sub-title ‘a study of magic in its religious foundations in Sanskrit texts, with a particular attention to a fragment on Viṣṇu’s Māyā preserved in Bali’ – a very interesting study.
The TBhS, chapter 10 (pp. 359–366) gives the rules for mantravicāra to be practised with different diagrams for the ‘six [magical] acts’.
54 The KT 15.97 excludes thus kūṭamantras, these being however, according to the tantra’s commentary, mantras consisting of a group of consonants. The same tantra (15.97–99) excludes also different mantras, such as those given by a woman, received in a dream, or those originating from the four Kula traditions (caturāmnāyaja) – which would exclude at least the mūlamantras of all the Kaula deities – and so forth. In this case, as in other ones, exceptions are invoked to blunt the edge of the general rule.
55 These mantras are among those excluded by the KT, as said in the preceding note.
56 For instance in the Sanatkumārasaṃhitā, Indrarātra, 2.
1 This is an English version, corrected, emended or when necessary updated of the paper japa, published in the BEFEO, LXXVI, 1967, pp. 117–164.
2 The Amarakośa 2.7,47 says: svādhyāyaḥ syaj japaḥ, underlining thus either the role of japa in the svādhyāya or the fact that all japa is a sacrifice of the word, and thus, like the svādhyāya a mahāyajña, a very important ritual action.
3 The Brāhmaṇas relate silence (tuṣṇīm), mumbling (jalpa), speaking in a low voice (upāmśu) to the indefinite or illimited (anirukta): ŚBr.5.4.4,13). The mumbling or murmur (jalpa) is associated with life (AiTBr. 2.39,1. See L. Silburn, Instant et cause, Paris, 1995.
4 According to the LT (39.35), the vācikajapa is to be used for minor rituals; the upāmṣu for rituals which bring success or power (siddhikarmāni), the mental (mānasa) one, for rites which bring both success and liberation. The japa done by dhyāna (meditation or visualisation of a deity?) is for achieving success in all cases (sarvasiddhi).
The three sorts of japa are sometimes related to the three levels of vāc: paśyantī, madhyamā and vaikharī – thus Kṣemarāja ad SvT 2.146–7. Or the commentary of Kṣemarāja on SvT 2. 146–7:
ātma na sr̥ṇuteyaṃ tu mānaso ‘sau prakīrtitaḥ
ātmanā śruyate yas tu tampāṃśuṃ vijānate/146/
pare ṣr̥ṇvanti yaṃ devi saśabdaḥ sa udāhr̥taḥ/
where he explains:
manaso madhyamāyam vāci/upāmśusaśabdau tu
sūkṣmasthūlaprayatnayāṃ vaikharyam.
5 These aspects of japa in the MBh are studied by V.M. Bedekar, ‘The place of japa in the Mokṣadharmaparvan (MBh XII. 189–193) and the Yogasūtras’, ABORI, vol. XLIV, 1963 (pp. 63–74). On the relationship of japa with Yoga, see J.W. Hauer, Der Yoga (Stuttgart, 1958), pp. 198–199.
6 hotr̥ japaṃ japati retas tat siñcatyupāmśu. japaty upāṃśviva vai retasaḥ siktiḥ. The same text (2. 31–32), regarding the recitation of the silent praise, tūṣṇīṃ śaṃsa, which consists in the japa of bhur bhuvaḥ svar, considered as the essence of the sacrifice, parallels it, too, with the emission of semen. See L. de La Vallée Poussin, Etudes et matériaux (London, 1898), pp. 120–123.
7 On Hindu conceptions concerning sexual energy, its spending or its storing, see Sudhir Kakar, The Inner World. A Psycho-analytic Study of Childhood and Society in India (Delhi, 1981). See too, A. Padoux, ‘Le monde hindou et le sexe’, Cahiers Internationaux de Sociologie, vol. LXXVI, 1984.
8 I refer to Alexis Sanderson, ‘The Śaiva Age – The Rise and Dominance of Śaivism during the Early Medieval Period’, in Shingo Einoo, ed., Genesis and Development of Tantrism (Tokyo: Institute of Oriental Culture, University of Tokyo, 2009, pp. 41–350).
9 A unity that reflects the unity of the Hindu world. Buddhist ṭantric practices, too, are marked by the same representations (on the creative efficacy of consciousness or on the image of the body notably).
10 ‘It has come to designate, in classical India, all types of prayer’, see L. Renou, ‘La valeur du silence dans le culte védique’, JAOS, 1979, pp. 11–18.
11 Hélène Brunner, who read a first draft of this chapter and made a number of useful suggestions, remarked that it would have probably been more rational, and clearer, for me to distinguish between the cases where the worship is the main portion and raison d’être of a rite, japa being accessory, and those where the japa is the essential part, the worship being accessory. She also drew my attention to the fact that it is in kāmya rites – especially those performed by the sādhaka – that japa takes the larger place, which is true, see note 16.
12 As, for instance, for Gaṇeśa: OṂ HRĪṂ ŚRĪṂ ṄLAUṂ GAṂ Gaṇapataye vara varda sarvagaṇaṃ me vaśa ānaya THAḤ THAḤ.
13 akārādi kṣakārāntam aksaramālā vidur budhāḥ (Japalakṣaṇam – ms. no. 545, Gov. Orl. Library, Madras). The Nityotsava (NU), p. 127, enumerates thus the rosaries made of the varṇas of the mātr̥kā ‘from the letter A to the letter KṢA’ and those made of rudrākṣa, pearls, rubies, crystal and so forth. The japa of the varṇamālā is prescribed in a number of texts. A Śaiva Upaniṣad, the Akṣamālikopaniṣad, describes this japa, which is done with a rosary and includes various mental visualisations.
14 See Pheṭkāriṇī (PheṭkT) 18.1–8:
bhūtalipyā pūṭīkr̥tyā yo mantram bhajate naraḥ /
kramotkramācchatāvr̥ttyā tasya siddho bhaven mantraḥ //1//
The following ślokas describe this action and its effects. This technique is also prescribed by the KT 15.16. On saṃpuṭīkaraṇa, see here, chapter 7, note 5.
15 mātr̥kājapamātreṇa mantrāṇāṃ koṭikoṭayaḥ/japitāḥ syur na sandaho yataḥ sarvaṃ tadudbhavam //18// – KT 15.18).
16 See, for instance, the japa of the gāyatrī among Brahmins of Gujerat described by L. Sinclair Stevenson, The Rites of the Twice-Born (New Delhi: Orl. Books Reprint Corporation, 1971), p. 223. Or, more generally, R.B.Ś. Chandra Vidyarnava, The Daily Practice of the Hindus, chapter IV, Sandhyā,, especially pp. 39–42.
17 Various prescriptions on favourable periods for japa are to be found in TBhS, pp. 316–317, where several texts are quoted. The JayS 14.188–189 prescribes different periods according to the nature (sātvika, etc.) of the adept. See also AgPur 292.31ff.: mantrasya devatā tāvat tithivāreṣu vai japet:/ etc.
There is also a rule in several texts (AgPur, ŚT, TRT, etc.) saying that japa is fruitless when done while the mantra ‘sleeps’: svāpakāle tu mantrasya japo na phalapradaḥ, says the Rudrayāmala quoted in TBhS, p. 324. But the times when a mantra is ‘asleep’, or ‘awake’ are not to do with the time of clocks. For the AgP (292.8) they are linked with the flow of prāṇa in the nāḋis idā and piṅgalā (see Padoux ‘On Mantras and Mantric Practices in the Agni-Purāṇa’, Purāṇa, vol. XX/1, 1978, p. 59). Other texts say that the ‘sleeping’ state of the mantra is the effect of a doṣa due to its phonetic structure – so ŚT 2.84: trivarṇo haṃsahīno yaḥ suśuptaḥ sa udīritaḥ – this defect is cured, the ŚT says, by showing the yonimudrā together with the udāna prāṇa. The commentary on this śloka (ŚT 2.110, vol. 1, p. 94), however, says that the correction of the defect is necessary only in kāmya rites. This point would tend to justify Hélène Brunner’s suggestion I mentioned previously (note 10) that insofar as japa is a manipulation of the power of the word for particular ends, japa would especially be used by sādhakas rather than by liberation-seekers (mumukṣu).
18 Infra, p. 42.
19 This is also called pūrvaseva, which designates more generally the preliminary propitiation of a deity by the sādhaka to obtain supernatural powers – see the entry on that word in TAK3.
20 I will come back on this later (see pp. 34–35). On the mantrasādhana, see Hélène Brunner’s study, ‘Le sādhaka, personnage oublié du śivaïsme du sud’, J.As., 1975, pp. 411–443, where, for the choice of the place, the author refers notably to the Mr̥g Cp. 1.95ff.
21 KT 15.22–30. Similar prescriptions are to be found in the PRKS 3.2.35–3/1 (p. 544), or in the TBhS, p. 317, which quotes three different texts; or in the Makuṭāgama 2.766–768 (information H. Brunner).
22 See A. Sanderson, ‘Power and Purity among the Brahmins of Kashmir’, in M. Carrithers, et al., eds., The Category of the Person (Cambridge: Cambridge University Press, 1985), pp. 200–202.
23 See, for instance, the prescriptions of the JRY 2.11 for the sādhana of the goddess Mahājhaṇkariṇī: the japa is done at night, in a cremation ground, a temple to the Mothers or a place for ‘Heroes’ (vīrasthāna), while wearing black clothes, etc.
24 Before a ritual is performed it is necessary to enclose the ritual area by means of a rite named digbandhana, which both delimitates and protects it from dangers coming from outside. On such a rite, see, for instance, TBhS pp. 317–348.
More generally, on the question of the protection of the ritual area against dangers and impurity felt as always menacing – on the Hindu preoccupation with boundaries, see D. Shulman, ‘The enemy within: idealism & dissent in South Indian Hinduism’, in S.N. Eisenstadt et al., Orthodoxy, Heterodoxy and Dissent in India (Berlin: Mouton, 1984), pp. 11ff.
25 In principle, one must not be lying nor standing – see KT15.032–33, or the MatP Kp. 11.12b–13 where other references are given (information H. Brunner).
26 See Mr̥g Cp. 75b: sādhakaḥ. sādhyaveṣadharaḥ. The same text, Kp. 3.1–45b–47, mentions the colours of the different aspects of Ṣiva (inf. Brunner). All worship (as well as many other rites) are to be performed wearing the proper sort of clothing. Usually it must be pure and clean; but in Tantric context it may, on the contrary, be chosen so as to mark the transgression: rags, black material for the Kāpalikās, transvestism in goddess cults, and so forth.
27 Such pavitras are usually rings of sacred herbs bound around the finger. On the subject, see the SP, vol. 2.
28 On all these points, see Rāghavabhaṭt .a’s commentary on ŚT 16.54–56 (vol. 2, pp. 656ff., which is on puraścaraṇa, and quotes a number of texts. Detailed prescriptions are also found in recent works, which shows that these instructions are still valid: see, for instance, Karanidan Sethiya, Mantra-vidyā, Calcutta, sam. 2033 (in Hindi), pp. 10–16 – the directions given concern japa in general. For the author, the colour of the clothes to be worn must be in accordance with the bhāva of the deity. Sewed garments are often prohibited – which is a general rule among observant Brahmins, for all rituals.
29 The Brahmayāmalatantra is as yet (2009) unpublished, It has been studied and partly edited by Shaman Hatley (thesis, 2007).
Bhrūkuṭi, contracted brow, is a facial expression typical of wrathful deities – Bhairava, here. Aṭṭahāsa is the laugh of Bhairava or other wrathful deities (such, for instance, as the corpse-like Sadāśiva imagined by the yogin as lying at the base of the triśūla on top of which are the three supreme goddesses of the Trika – Parā, Parāparā and Aparā.
30 This still unpublished tantra is being studied and is in the course of edition by Judit Törzsök.
31 On these nyāsas see the next section.
32 NU, p. 93; Prāṇatoṣiṇī, p. 421, quoting the Sarrasvatītantra, which prescribes a preliminary purification of the mouth, mukhaśodhana. The GT prescribes the utterance of the setu OṂ before and after the japa. Other prescriptions in the JS, chapter 14.
33 NU pp. 93 and 94.
34 See for instance the LT 28.47–50 already quoted which prescribes alternate japas and meditation.
35 ŚT 16.56, comm. (pp. 655–6). For Kr̥ṣṇa rites, see R.V. Joshi, Le rituel de la dévotion kr̥ṣṇaïte (Pondichéry, 1959). Many informations given there come from the Haribhaktivilāsa, an important ritual manual. See too the Nāradapañcarātra – Ṣrī Nārada Pañcaratanam (Sacred Books of the Hindus, 23, reprint, New York 1974).
36 The comm. on Mr̥g, Cp 1.76 prescribes audible, murmured or mental japa depending as to whether the mantra to be mastered by mantrasādhana is of an inferior, middling, or superior sort (without saying what are these three sorts of mantras).
37 Thus NU, pp. 129–130, quoting the Svachandatantrasāra: japas tu ṣaḋvidhaḥ vācika mānasa yogavācika yogamānasika vaṇmānasikayaugikaḥ – the four last sorts corresponding to the association of the utterance of the mantra with yogic practices. On japa, prāṇa and yoga, see further down. The Makuṭāgama 2.760 lists japas in descending order as follows: mānasa, kaṇṭhoṣtha (?), upāṃśu and vācika.
TĀ 29. 82–88, quoting the Yogasaṃcāratantra, describes different sorts of japa which are yogic practices involving the movement of prāṇa.
38 ‘Hasty japa brings about sickness and slow utterance destroys wealth’ writes a contemporary author: introduction to an edition of the Mantramahodadhi (Mud) (Delhi, 1981)
39 YS 1.28: tajjapastadarthabhāvanam ‘Its [sil. of the mantra] repetition and concentration on its meaning [should be made]’.
40 One may well ask oneself, in such cases, whether the mind of the practitioner is fastened on this mental image, or on the sound of the mantra he recites. Perhaps, the sound forms, as it were, a basis whence the visual image arises.
41 On this subject, see the studies of J.F. Staal (for whom rites are without meaning), The Science of Ritual (Poona, 1982), and Ritual and Mantras: Rules without Meaning (Indian edition, Delhi, 1996).
42 See YH 3.191, and Dī, p. 332. The japakāla may last a fraction of a second: there is no real duration. But such japas are in fact a mental attitude, a bhāvanā, of the practitioner. Many such instances are to be found – see, for instance, the Kulamūlaratnapañcakāvatāra of the Kubjikā tradition, 3.17.
43 In the Mahādevayajña ritual of Tanjore, for instance (in 1951, as reported by C.G. Diehl in Instrument and Purpose, Lund, 1956), were performed a hundred thousand repetitions of the thousand Names of the Goddess, a hundred million recitations of the śrīvidyā, etc.
44 PRKS, p. 550. The form or the material of the seat, the clothes, the bodily posture, etc., vary according to the context and circumstances of the japa. See, for instance, MatP Kp 11.12ff. (vol. 2, p. 198). In all cases one is to follow the prescriptions given for the ritual which includes the japa.
45 One could say here that insofar as we may consider mantras to be performative utterances in the sense of Austin and Searle, the circumstances mentioned previously in which the japa is to be made to produce the desired effect are merely an intensely meticulous case of the necessity of the appropriate social conditions of utterance for it to be effective as defined by these two authors. On performative utterances, one may refer to J.L. Austin, How to Do Things with Words (Oxford, 1975), and to J.R. Searle, Speech Acts (Cambridge, 1980).
46 Śākta texts sometimes prescribe counting on eight phalanxes only, which results in sixty-four (8 × 8), an important number in Tantra: there are sixty-four Bhairava tantras, sixty-four Yoginīs, and so forth. See for instance Yoginītantra 2.30, Another method (on nine phalanxes), GT 18.11–12.
47 The JayS 14.72–76 (meror taṇghananiṣedhaḥ) explains how to do this. For japa of the varṇamālā,, KṢA is considered as the Meru, but it can be included in the count: see Mantrakaumudī of Devanātha Thakura (Darbhanga, 1960), 14: anulomavilomais tair binduvāṇmātr̥kākṣaraiḥ/kṣamerukaiḥ sāṣṭvargaiḥ kl̥iptayā varṇamālayā //13//.
If one uses the hand for counting (karamālā), the second of the middle finger (at least according to some texts) is to play the role of the Meru. But the counting is not done going forward and backward (see Yoginītantra 2.40: anulomavilomabhyāṃ sarvamālāsu saṃjapet/kevalaṃcānulomena prajapet karamālayā). For the JayS 14.72ff., the rosary is always to be counted in the same way, a particular gesture avoiding the overstepping of the Meru.
When one does japa of the phonemes (varṇamālā), from A to KṢA, one recites it twice, anuloma then vilomena, the first phonemes of each of the eight varga (A, KA, TA, etc.) so as to reach a total of 108. According to the JayS, the akṣamālā has 108, 54 or 27 beads.
48 This is a case of the rule prevalent in the Hindu world that the index is never to be used except in destructive or malignant rites. See GT 18.27–30 which prescribes the use of different fingers for different aims. Prescriptions of the same sort in the Śivārcaṇacandrikā quoted in SP1, pp. 216–217: one must use different fingers depending on the the japa being audible, whispered, or mental.
49 PRKS, p. 551: vastreṇācchādya tu karaṃ dakṣiṇaṃ yaḥ sadā japet.
50 See YH 3.189b: ‘the japa once thus accomplished, it must be offered in the left hand of the Goddess’, evaṃkr̥tvā japaṃ devyā vāmahaste nivedayet.
51 If a new rosary breaks, one is to repeat a thousand times the Aghoramantra, but merely a hundred times if the rosary is an old one (see prāyaścittavidhiḥ, 48, SP2, pp. 260– 261). The breaking of the rosary is karmachidrakaraṇam: it causes the destruction of the [good] karman (JS 14.93).
52 OṂHŪṂ VAUṂ Viśvakṣeṇāya namaḥ.
53 Pp. 325–329 of the TBhS are on puraścaraṇa., enumerating or describing its rites. See too Rāgahvabhaṭṭa’s commentary on ŚT 15.54 (vol. 2, pp. 656ff.) which quotes many texts; or GT. 18. For the vaiṣṇava rite, see the Nāradapañcarāttra, III, chapters 10 to 14. Kane, History of Dharmaśāstra, vol. 5/2, pp. 1008–1112 are on puraścaraṇa, with many textual references. See also H. Brunner’s study, ‘Le sādhaka’ quoted in note 13 of chapter 2.
54 KT 15.18, the pūjā of this list is the threefold daily sandhya – a nitya ritual.
55 See Hélène Brunner’s (French) translation, and her notes, pp. 282–284 of the study quoted in notes 16 and 44.
56 The ŚT 18.14 (vol. 2, p. 716) prescribes 24 million (tattvalakṣaṃ japan mantram) repetitions of the mūlamantra of Śiva, followed by 24,000 oblations, which implies that the ritual lasts several days or months. Manuals of ritual discuss such details and the time necessary for such practices.
57 See thus the Nāradapañcarātra, III. chapter 14, śl. 15ff., where the japa of two mantras of Kr̥ṣṇa protects children against demons (15), keeps away harmful astral influences (19), gives long life (28), brings rain (45), protects from illnesses and from sudden death (52), etc. This japa is to be accompanied by offerings in the fire whilst meditating and visualising the child Kr̥ṣṇā in his ‘games’ (līlā).
58 Cf. GT 18.68: agnir eva mukhaṃ devyās tasmād agnau hutet sadā. Those are ancient notions.
59 Kane (op. cit., p. 1109) says thus that, according to the Kaulāvalinirṇaya, a rite of japa is to be done during the night in a cremation ground with a corpse. The TBhS (p. 334), quoting the Svātantratantra, describes another nocturnal rite where the japa accompanies sexual ritual practices.
60 The same prescription (but without the mudrā) is found in the Mr̥g. Kp. 4, 5 (p. 42, where Bhaṭṭanārāṇayakaṇṭha’s commentary is very explicit).
61 This is at least how the rite appears to me. H. Brunner, who knows the subject much better than I do, feels that I stretch unduly the meaning of this action. For her, the flower symbolises the whole japa, this symbol being what is given to the deity (see Mr̥g’s French translation, p. 55, note 2).
62 But a Tantric pūjā is always seen as a process of identification with the deity that is worshipped, an identification that is ritually achieved (or played out) several times during the cult.
63 On these elements, see here chapter 4 on nyāsa.
64 devīnāṃ vāmahaste tu devānāṃ dakṣiṇe ‘rpayet, says the Rudrayāmala quoted in the Tāriṇīpārijata, p. 94: this eighteenth-century work gives a number of details on the japa (japamālā), of the phonemes of the alphabet, pp. 87–89.
65 The LT 40.18 says (the goddess is speaking): ‘having completed the japa according to the prescribed method, he should give it over to Me and imagine the śakti in the form of japa as being in my mouth’. Here too the deity absorbs the power of the word when the utterance of the japa ends. The officiant is now to ring a bell identified with Sarasvatī, the goddess of speech, which will cause the mantras to come and thus contribute to the efficacy of the recitation. There follows a ritual worship which will ensure the mantrin’s mastery of the mantra or the fruitful result of its use (mantrasiddhiṃ nigacchati – śl. 29).
66 See for instance Daily Practice of the Hindus, chapter 9, ‘Sandhyā of the Yajurvedins’, section 23, japanivedana: having recited the gāyatrī 108, 28 or 10 times, the officiant, seated in front of the image of the deity, pours some water in the palm of his right hand while saying anena (daśa, aṣṭaviṃśati, etc.) saṃkhyayā japam. bhagavan (Brahmā Visṇu, etc.) svarūpi sevitā priyatāṃ namaḥ, and pours that water in the hand of the devatā (ibid., p. 109).
On the other hand, the private cult that the temple officiant performs at home (or in the temple) for himself, where the japanivedana we have seen takes place, is entirely Tantric.
A mental nivedana linked to the movement of prāṇa, at the end of the japa, is prescribed in SvT 2.136–143.
We may, apropos of the practices described above, underline once more the fact that Tantric practices, far from contradicting the Brahmanic tradition (even if it often runs counter to it), takes place within this tradition: Tantra, I believe, brings to light the inner virtualities of Brahmanism.
67 Here is, as an example, how a small modern manual (Śrīkāīīnityārcaṇa, Prayāg, Kalyan Mandir, saṃ 2030, p. 89) describes (I have shortened a little) the japa at the end of Kālī’s pūjā: after the preliminary rites, pay homage (tarpaṇa and pūjā) to the rosary with the mantra Śrīdakṣiṇakālikā-akṣamālāṃ tarpayāmi pūjāyāmi namaḥ; then pray to the rosary to give success (me siddhidā bhava). One holds it then with the right hand and says another prayer to remove all obstacles. Having saluted the rosary, one will place it on the middle phalanx of the middle finger and touch it with the thumb saying OṂ, which is done while ‘meditating’ (dhyāna) the form of the goddess. Then one will pray to the rosary, asking it to grant happiness and success. Having held it to his head while reciting as before the setu and mahāsetu, the adept or devotee will practise a prāṇāyāma. He then does a r̥ṣyādinyāsa and a karāṇganyāsa. Then the japa itself is to be done. Once finished, it is to be offered, together with a special arghya and amr̥ta, in the left hand of the goddess, while saying OṂ guhyātiguhyagoptri tvaṃ gr̥hanas matkr̥taṃ japaṃ. siddhir bhavatu me devi tvatprasādānan maheśvari. idaṃ japaṃ śrīdakṣiṇakālikādevyā samarpayāmi svāhā. Then one salutes her making the yonimudrā and saying HRĪṂ siddhayai namaḥ. After which the rosary is to be kept in a secret place. The japa is to be followed by a homa.
The TĀ 29, describing a pūja according to the Mādhavakula and quoting the Yogasaṃcāratantra, prescribes a japa of a 100,000 to 1,200,000 repetitions, linked to the ascent of kuṇḋalinī; it is said to be a murmuring (saṃjalpa) whose essential nature is an awareness of the subtle phonic vibration (nādāmarśasvarūpiṇī).
68 In the case of the japa (as part of the mantrasādhana) prescribed in the NṢA 5, according to the number of repetitions (from 100,000 – one lakh – to 900,000), the mantrin would be freed from the sins of his previous lives, would be granted the eight ‘great powers’ (mahāsiddhi), or, for nine lakhs, the fusion with the deity (ibid., śl.5–6).
69 The fact that śānti, a favourable rite, is among the six magical actions (and is even often quoted first), as well as the fact that these acts are usually described in mythical and religious texts, show that the distinction religion/magic which I make here is problematic and, in fact, quite difficult to apply in the Indian context. This would also raise the more technical problem of the distinction between kāmya rites (that of śānti notably) and nityā rites – on which see H. Brunner’s observations in her seminars at the Ecole Pratique des Hautes Etudes, Sciences Religieuses, in 1984.
70 See also Vīṇaśikhatantra (ed. Goudriaan, Delhi, 1985), śl. 224ff.
śāntipuṣṭivaśākr̥tyucāṭana māraṇe /
svadhāsvahāvaṣaṭ hūṃ ca vauṣaṭphaṭ yojayet kramāt //
See too the explanatory note 62 in SP3, p. 166, or ibid., pp. 532–533, on astrābhiṣeka (ritual sprinkling). In the māraṇa, etc. rites, one utters the mantra without bindu or nāda.
72 One finds prescriptions of the same sort in the Śeṣasaṃhitā (14.58–64,a,d 15, pp. 79–83) in a passage concerning the ṣaṭkarmāṇi. Concerning the use of fingers, that of the index for destructive rites is normal (see above). In fact, prescriptions vary largely according to texts. The relationship between some physical aspects of japa and the nature of the rite does not appear to us prima facie to be logically grounded – though, admittedly, a more precise research on the subject may uncover symbolic reasons for this situation.
On the magical uses of mantras in contemporary India, one may refer to C.G. Diehl’s study, Instrument and Purpose, previously quoted.
73 See the observations of Hélène Brunner in SP3.
74 The omissions or errors repaired by prāyaścitta are, so to speak, technical accidents during the utterance. They are not the procedural errors or omissions – the ‘infelicities’ – which, in Austin’s then in Searle’s theory, make the speech acts ineffective. (On mantras as illocutionary or perlocutionary acts, see A.C.S. McDermott, ‘Towards a pragmatics of mantra recitation’, Journal of Indian Philosophy, vol. 3, 1975, pp. 283–298.)
75 See H. Brunner’s study ‘Le sādhaka.’, JAs, 1975, previously quoted.
76 Sāttvatasaṃhitā, with a commentary by Alasinga Bhatta, edited by V.V. Dvivedi (Varanasi, 1982).
77 Thus the brahmavidyā recited in the ear of a dying person, the dīkṣā being in that case limited to the repetition of a mantra, according to the TĀ 19. 31–33 – the recitation, in that case, is a mere ‘reading’ ((paṭha) of the mantra.
78 See, for instance, SP3, note 54a, p. 221, where is quoted the Mr̥g, Kp 7, 98–110, prescribing a japa of the Svapnamāṇavakamantra, an invocation of the god of dreams (necessary to be known for the dīkṣā).
79 In the funerals, antyeṣṭi, for instance (see SP3, pp. 372–3 and 610), but it is in fact a prāyaṣcitta; or, in the cult of ancestors, the ṣrāddha (ibid, pp. 656–657. The MatP Cp 10.73 prescribes 108 repetitions of the Bahurūpamantra to purify those who bear the corpse of an ascetic (yātin) to the place where he will be buried.
80 On this point, see Milton Singer, ‘The Radhâ-Krishna bhajanas of Madras city’ in M. Singer, ed., Krishna: Myths, Rites and Attitudes (Honolulu: East-West Press Center, 1966). See too the observations of Hélène Brunner in the Supplément of vol. 3, ‘Rites et Fêtes’, Fr. L’Hernault & M-L. Reiniche, eds. of the study carried out and published by The Ecole Française d’Extrême Orient, Paris, 1990–1999, Tiruvannamalai, Un lieu saint sivaïte du Sud de l’Inde.
81 See pp. 377–392 of that translation; a chart is given on p. 378 of the ascent of the śrīvidyā along the suṣumnā.
82 On these, see the very detailed notes of H. Brunner in SP3, pp. 358–396.
83 Infra, p. 46.
84 Commenting on the passage of the YH 3 on the japa of the avasthās and on the śūnyas, Amr̥tānanda uses the term bhāvanā to describe this process. On bhāvanā, see F. Chenet, ‘Bhāvanā et la créativité de la conscience’, Numen, vol. XXXIV, 1987, pp. 45–96.
In an analogous though somewhat different perspective, the TRT 38.18, apropos of the cult of the Yoginīs in the śrīcakra, mentions three different ways for identifying with the deity: nyāsa, which makes the body of the adept similar to that of the deity, japa, which creates this similarity (japas tanmayatārūpabhāvanam), and homa, as a symbolical offering of the universe which dissolves the multiplicity of the created world into the unity of the Self.
85 This sort of japa is sometimes described as a remembrance (smaraṇa). Thus YH 3.184, on the japa of the avasthās, or the Varivasyārahasya of Bhāskararāya, 52, on the japa of the avasthās, śūnyas and viṣuvas. This is due to the influence on these authors of the Pratyabhijñā system for which smaraṇa is a means for the realisation of the Absolute since, by linking the past to the present, it focuses the attention on the supreme reality underlying all phenomena.
86 This information I owe to Alexis Sanderson.
87 F.D. Lessing and A. Wayman, op. cit., p. 187.
On another analogous Buddhist practice, see Stephen Beyer, The Cult of Tārā (Berkeley, 1973), p. 450–452, ‘Reciting the mantra’. Simpler is the japa of chapter 3 of the Mahāvairocanasūtra (ed. Tajima, pp. 117ff.) where a meditative recitation (bunsho manju) is done by contemplating in one’s heart the shape of the letters of the mantra: one sees there the importance of the written sign in Sino-Japanese culture.
88 See Padoux, Vāc.
89 The Yogīśvarīmata, also called Siddhayogeśvarīmata, Siddhamata, Siddhatantra, etc., is often quoted in the TĀ. Its doctrine is said to be expounded in the MVT. A critical edition of selected chapters of this tantra, with an annotated translation, was done by Judit Törzsök in an unpublished habilitation thesis (Paris, 2010). However, few of the quotations of this tantra given in the TĀ are to be found in this edition, which is perhaps that of a shorter version.
The Yoginīkaula is quoted only once in the TĀ: is it the same text as the Yoginītantra mentioned in the JRY?
90 He quotes the following śloka:
japet tu prānasāmyena tataḥ siddhir bhaved dhruvam /
nānyathā siddhim āpnoti hāsyam āpnoti sundari // (vol. 4/2, p. 30).
91 Here is Jayaratha’s commentary: tattadekapiṇḋādyātmakamantrarūpatayā bahir ullasantī varṇakuṇḋalinyākhyā parameśvarī śaktir yadi nāma prāṇasamā prāṇasāmyenodayam iyāt tadunmanā śvaikātmyena prasphured ityarthaḥ (p. 32).
92 This at least is how one usually translates haṃsa, this bird being in fact a duck of sorts.
93 See for instance, SvT 7.54–55 or VBh 155: ‘21.600 time by day and night this japa is prescribed as that of the Goddess’. See too TĀ 7.47–50, where various divisions of the mantra and of its utterance all result in 21,600. This number of breaths per twenty-four hours is in fact perfectly real.
94 Thus SvT, VBh, TĀ, ŚSV, such Upaniṣads as Dhyānabindu, Mahāvākya, Yogaśikhā, and so forth. In the Brahmavidyā Up., the japa of haṃsa is combined with that of OṂ and is associated to the ascent of kuṇḋalinī
95 Gāyatrīs can be composed on the model of the savitrī for all divinities.
96 A Tamil author who may have lived in the sixteenth century. This work was edited and translated in French and commented on by Tara Michael (Pondichéry: Institut Français d’Indologie, 1975).
97 Here is SvT 7.59:
nāsyoccarāyitā kaścit pratihanta na vidyate/
svayam uccārate haṃsaḥ prāṇinām urasi sthitaḥ//
The ślokas of VBh run as follows:
sakāreṇa bahir yātihakāreṇa viṣet punaḥ/
haṃsahaṃsety amuṃ mantraṃ jīvo japaḥ nityaśaḥ//
ṣaṭśatāni divā rātrau sahasrāṇy ekaviṃśatiḥ/
japo devyāḥ samuddiṣṭaḥ sulabho durlabho jaḋaiḥ//
They can be translated as follows:
With the letter sa [breath] flows out; with the letter ha it flows back.
Thus the living being recites this mantra haṃsahaṃsa.
Twenty-one thousand times, night and day, is this japa of the supreme goddess,
it is very easy [but] difficult for the ignorant.
98 See, for instance, the Gorakṣaśataka quoted by Briggs, in Goraknāth and the Kanphata Yogis (Delhi: Motilal Banarsidass, 1982), pp. 292–293; or the Yogaviṣaya, 29, p. 47 of Kalyan Mallik’s Siddhasiddhāntapaddhati & Other Works of Nāth Yogis (Poona: Oriental Bookhouse, 1954; or the Yogamārtanda, 33ff. (ibid., pp. 58–59). The ajapājapa is still in current use even in non-yogic or ascetic milieus. An analogous practice is to be found (with so’ham. instead of haṃsa) in the Haṃsopaniṣad, 10–13, the Brahvidyopaniṣad, 78–79, among the Sants, etc.
99 See the texts quoted by S.B. Dasgupta in An Introduction to Tantric Buddhism (University of Calcutta, 1950), pp. 122–125, and in Obscure Religious Cults (already quoted here), pp. 96–99.
samyendriyasañcaraṃ proccaren nādam āntaram /
eṣaiva japaḥ prokto na tu bāhyajapo japaḥ //
This stanza comes perhaps from the Saṇketapaddhati, an important work of the Śrīvidyā known only by quotations.
101 Except probably insofar as the inner perception of the nāda can be actually perceptible as some sort of subtle inner phonic vibration. Several works describe different sorts of nāda accessible to yogins, some of which appear as perceptible. There is thus a division into sixty-seven nādas, eleven of which are ‘unstruck’ (anāhata) sound: see Vidyānanda, Artharatnāvalīūī ad NṢA 1.12. The subtle form of nāda is often compared to the end of the vibration of a bell which becomes ever more fine untill it disappears altogether in utter silence.
bhūyo bhūyaḥ pare bhāve bhāvanā bhāvyate hi yā /
japaḥ so’tra svayaṃ nādo mantrātmā japya īdr̥śaḥ //
103 The I is aham, which written a+ h +ṃ symbolises the totality of the cosmic manifestation which, going from the first (A) to the last (HA) letter of the Sanskrit alphabet, is brought together into a unique dot, the origin of everything, the bindu Ṃ.
A and Ha are sometimes taken as symbolising the manifestation and the resorption of the cosmos; thus Ksemarāja ad SvT7, 29.
104 This tantra, also named Triśirastantra, Trśiromata, etc., is often quoted by Abhinavagupta. It is known only by quotations. It apparently expounds the conception of the supreme godhead as made of the three goddesses Parā, Parāparā and Aparā.
kalmaṣakṣīmanasā smr̥timātranirodhanāt /
dhyāyate paramaṃ dhyaṃ gamāgapada sthitam /89/
paraṃśivaṃ tu vrajati bhairavākhyaṃ japaḋ api /
tatsvarū paṃjapaḥ prokto bhāvābhāvapadācyutaḥ //90//
106 Parāmarśa, this essential notion of the Trika, is very difficult to translate; see the entry on that word in TAK, Vol. 3.
107 This text seems to be known only through quotations by Kṣemarāja and a few other authors. It is likely to be a Krama text.
108 mahāmantrātmakākr̥takāhaṃvimarśārūḋhasya yad yad ālāpādi tat tad asya savātmadevatāvimarśānavaratāvartanātmā japo jāyate.
109 For Bhāskara (ŚSVārttika), japa appears in this case as an enunciation resting on energy (vīrya) and leading to identification with the supreme Being. According to him, it can be of four sorts: ‘of energy’ (śākta), identifying the soul with the Self, ‘without parts’ (niṣkalā), which seems to be the utterance of OṂ, that of the haṃsa ‘made of the klaās of nāda – nādātmika – and finally the japa of the living being (jīva), or of the body (paudgala) which is probably the movement of respiratory breath. This commentary is earlier than Ksemarāja’s, and probably nearer to the spirit of the ŚS.
110 We might say of haṃsa with Lilian Silburn: ‘All human beings recite it perpetually, without their knowing it, whether awake or asleep. But the jñānin, different in that from the ignorant, is aware of this haṃsa; he enjoys the presence of Śiva united with his energies, powerful, omniscient’ (my translation from p. 178 of L. Silburn’s French translation of the Śivasūtras with Kṣemarāja’s Vimarśinī).
akr̥trimataddhr̥dayārūḋho yat kiṃcid ācaret /
prānyād vā mr̥śate vāpi sa sarvo’sya japo mataḥ //
112 In the fourth chapter of the Tantrasāra, which gives a shortened version of TĀ 4, Abhinavagupta says that pure intuitive reasoning (sattarka), the pure gnosis (śuddhavidyā) which is the way towards the Absolute can be attained by sacrifice, by oblation in the fire (homa), by japa or by yoga, all of which he defines as an immersion in divine Consciousness. Of japa, he says: ‘it is an inner intense awareness of the fact that the supreme Reality exists by its own nature, independently from the existence of the diversity of the cognisable, for all that is then to appear is a consciousness which grasps and transcends all this dichotomy’ (ubhayātmakaparāmarśodaya).
matparaṃ nāsti tatrāpi jāpako’smi tadaikyataḥ /
tattvena japa ityaksamālayā diśasi kvacit //
114 When commenting on this stanza, Kṣemarāja adds: ‘You teach that this japa which in essence is an awareness of the Unsurpassed … is to be accompanied by the placing [on the body] of the letters of the alphabet’ (japaḥ … anuttaravimarśasāro … varṇalipinyāsena yuktyā śikṣayasi).
In fact, all japas are oriented towards an identification with the deity. This is obvious for the devotional japa, as we shall see. But even in the puraścaraṇa preliminary to kāmya rites, this identification is described as necessary: without it the japa is useless.
115 See, for instance, the YH 3.3 where Bhairava justifies the daily practice of the pūjā of the śrīcakra by saying: ‘It is performed perpetually by me’ – cakrapūjā ca sadā niṣ padyate mayā.
116 On this, one may refer to C.G. Diehl’s study, Instrument and Purpose, quoted here, note 43. See notably, pp. 267–324, on ‘Mantirikam’, where many prophylactic healing rites, rites for success and so forth are described, of which japa is an important element.
117 A difference which of course does exist up to a point in non-dualistic context, where the essential unity of the human being and the godhead, of the ātman and of the brahman, is effectively experienced at the end of the ascesis by the liberated in life. The Śivastotrāvalī is an instance of non-dualistic devotional literature.
118 Elsewhere he writes about japa in āṇavopāya: ‘shall we say that the repetition of a religious formula plays the same part in bringing about the liberation as does a lullaby in putting a child to sleep, a state of physical quiescence’. Further on, he compares the effect of the two other upāyas to autosuggestion (śāktopāya) or deep sleep (śāṃbhavopāya) where the loss of consciousness is complete – but sleep, in this case, is taken in the sense of the second avasthā, therefore as a state superior to wakefulness, which is not our perspective. The case is nevertheless worth considering.
119 See RV 10.63, 2: ‘For all your names are worthy of homage, worthy of praise, Gods, [your names] are also worthy of sacrifice’, viśvā hi vo namaśyāni vandyā devā utà yajṇiyāni vaḥ.
120 The same sort of japa is prescribed in Śaiva Upaniṣads (Atharvaśiras, Kaivalya, Jābali).
121 kiṃ japyenāmr̥tatvaṃ brūhīti/sa hovāca yajṇāvalkyaḥ śatarudriyeṇeti /
122 Cf. SkPur 1.1.14:
śiveti dvyakṣaṃ nāma vyāhariṣyanti ye janaḥ /
teṣāṃ svargaś ca mokṣaś ca bhaviṣyati na cānyathā //
123 Such practices seem to have existed earlier in Buddhism for the worship of Amitābhā. But both the ‘spirit’ and the actual practice of such a japa differ very much from those of the Hindu japa.
124 This according to Ch. Vaudeville, Kabir (Oxford: Clarendon Press, 1974). Curiously, the Haripāṭh of Jñāndev (8.2) says that the japa of Rama’s name is the ‘breath’ of Śiva: ātmā jo śivācā/rāmajapa: Ch. Vaudeville, L’Invocation. Le Haripāth de Dyāndev (Paris: EFEO, 1969), p. 64. See note 120.
125 See Milton Singer, ed. op. cit., note 80, ‘The Radha-Krishna bhajanas of Madras City’, especially pp. 118–121.
126 Adh.R, I.6.69; II.5.25; IV.1.78 – this reference is found in F. Whaling, The Rise of the Religious Significance of Rāma (Delhi: Motilal Banarsidass, 1988), p. 160.
127 Tulsidas, while affirming that no other name than Rama’s saves, admits however the efficacy of other names, notably those of Sītā and Bharata ‘for many are the names of the Lord’.
We may note that for Tulsidas Rāma is the mantra of Śiva who, at Kaśi (Varanasi), saves all beings by giving them this mantra. The Haripāth says (8.2) that the japa of Rāma’s name is the breath of Śiva – is this an allusion to the ajapājapa? ‘Hari, Hari, such is the mantra of Śiva’ says also the Haripātth quoted by Ch. Vaudeville, op.cit., p. 64–65.
128 Caitanyacaritāmr̥ta of Kr̥ṣṇadāsa, ādilīlā, 8.22, quoted in E.C. Dimock, The Place of the Hidden Moon (Chicago University Press, 1966), p. 226.
129 See A.K. Majumdar, Caitanya. His Life and Doctrine (Bombay: Bharatiya Vidya Bhavan, 1969), p. 226. Rūpagosvamin, in his Bhaktirasāmr̥tasindhu, 1.2,48, defines kīrtan as follows: nāmalīlāguṇādinām uccair bhāṣā tu kīrtanam, ‘the fact of saying or singing the name, the play and the virtues or qualities of the Lord’ (ibid., p. 143).
130 On the Naths as they exist now in India, the most recent study is that of Véronique Bouillier, Itinérance et vie monastique. Les ascètes Nāth Yogīs en Inde contemporaine (Paris: Editions de la Maison des Sciences de l’Homme, 2008).
131 The Haripath recommends also the silent japa (26.4): ‘Dñyandev has taken the vow of silence. He wears his rosary in his heart. This is where he invokes the Name of Hari’ (ibid., p. 146). This, however, does not prevent the Haripāth from recommending mainly the collective singing of the Name: ‘the [only] means of salvation is the company of the saints’ (5.4).
132 to karī tetulī pūjā/to kalpi to japi majhā /… /samādhi majhī // This quotation, like all cited here on the subject is taken from Ch. Vaudeville’s study of the Haripāth, quoted in note 124.
133 See Ch. Vaudeville, Kabir Granthāvali (Doha), (Pondichéry: Institut Français d’Indologie, 1957), pp. XVIII and 36.
134 Ibid., pp. XVII–XVIII. See also passages from the Gorakhbāni, the sayings of Goraknāth quoted by Ch. Vaudeville in the introduction to her Kabir (Oxford, 1974), p. 138: ‘Get hold of the śabda, O Avadhūt ! Get hold of the Śabda! The ‘stages’ (sthāna) are useless obstacles.’
135 This was, for instance, the belief of the Rāmanandins, whose mantra is OṂ Rāmāya namaḥ, or merely rāmāya namaḥ, for OṂ is considered as the equivalent of Rāma. On this and on Tantric elements in Kabir, see Ch. Vaudeville, op. cit., especially section V of the Introduction, ‘Tantric concepts in Kabir’s verses’.
136 One may mention here the persisting use of sumiran, mental japa (mānasajapa), in the contemporary sect of the Radhasoamis: a modern variant of a very ancient tradition. On the Radhasoamis, see H. H. Sahabji Maharaj Sir Anand Sarup, Yathārtha Prakāśa (Dayalbagh, Agra, Radhasoami Satsang Sabha, 1984).
137 On dhikr, see the interesting study of G.C. Anawati and L. Gardet, Mystique musulmane, aspects et tendances, expériences et techniques (Paris: Vrin, 1970), pp. 187– 258, which notes analogies between dhikr and japa.
138 According to Anawati and Gardet, op cit. p. 197, the existence of the practice of dhikr made in common is not attested before the twelfth century, and could have appeared under the influence of japa.
139 There are in Bengal Muslim Baùls, the Fakirs.
140 Their case, but also the whole question of the rapports between Islam and Hinduism, was studied by Sayid A.A. Rizvi, A History of Sufism in India (New Delhi: Munshiram Manoharlal, 2 vol., 1978–1983): vol. 1, ch. 6, ‘Interactions between Medieval Hindu Mystic traditions and Sufism’, vol. 2, ch. 8, ‘The Sufi Response to Hinduism’. Other more recent studies on the subject could be quoted; for instance Thomas Dahnhardt, Change and Continuity in Indian Sūfism (New Delhi: D.K. Printworld, 2002).
An interesting case of syncretism is that of the Ismaelis, see T. Kassam and Fr. Mallison, Ginans. Texts and Contexts. Essays on Ismaeli Hymns from South Asia in Honor of Zawahir Moir (New Delhi: Matrix Publishing, 2007); or Dominique Sila Khan, Conversions and Shifting Identities. Ramdev Pir and the Ismaelis of Rajasthan (New Delhi: Manohar, 1997).
141 A paragraph on the subject in the original French version of this article is not translated here.
142 But what is a prayer strictly speaking, if not what one uses to consider as such in the traditionally Christian Western world? Not a generally valid definition, therefore. Using the word prayer for japa cannot but lead to misunderstandings.
1 This is an English version by Gavin Flood of the article ‘nyāsa: l’imposition des mantras’, published in the BEFEO, vol. LXVII, 1980, pp. 62–102. The original text was, where necessary, corrected and updated by the author.
2 We also find the terms vi-ny-AS: ‘to put or place down in different places, spread out, distribute, arrange …; vinyāsa (masc.), putting or placing down … arrangement, disposition, order …, scattering, spreading out, putting together’ (Monier-Williams Dict. – we also need to retain the nuances of these senses).
3 The Brāhmaṇa and the Kalpasūtra testify to the ancient use of ritual touching: we meet them quite often in the Śatapathabrāhmaṇa, accompanied with mantra recitation (e.g. ŚB I.1,2, 10–11; I.2, 2, 4–11; III.2, 1, 4–6 etc.). The word used is abhi-MṚŚ, that is, to touch.
We note that it is the more recent smṛti texts which prescribe nyāsa, although certain authors condemn their use as non-vedic (avaidika): cf. Kane History of Dharmaśastra vol. 2, p. 319. From a certain period the use is general.
4 The nyāsas have not, to my knowledge, been the object of study in any European language except for Aghehananda Bharati ‘Symbolik der Berührung in der hinduistisch-buddhistischen Vorstellungswelt’, in Studium Generale, vol XVII, no. 10, 1964, Heidelberg, pp. 609–620. This article which places the practice in a general anthropological perspective is somewhat brief about the nyāsas themselves.
5 See, for example, ‘hand’ in Hasting’s Encyclopaedia of Religion and Ethics. For India we can observe that hand gestures have a protective value and are used for consecration or homage, as witnessed on ancient Buddhist monuments down to our own day. On this point see the study by Ph. Vogel The Sign of the Spread Hand or ‘five finger token’ (pañcāṇgulika) in Pali literature (Amsterdam, 1919). See also J. Auboyer Le Trône et son Symbolisme dans l’Inde ancienne, pp. 11–15.
6 Cf. my Vâc. The Concept of the Word in Selected Hindu Tantras (Albany: SUNY Press, 1990).
7 On the karanyāsa (or karāṇganyāsa), see later p. 70.
8 This is a natural connection insofar as mantras are a variety of speech-act, and that speech is bound to gesture. If this form of speech which is mantra can be regarded as relevant to language, it is naturally also bound to gesture since in many respects language is in gesture as much as in speech. Mantra and mudrā are systems of coordinated signs. It could be interesting to make a semiotic study of them, without losing sight of the fact that both are grounded in a deeper level of meaning to which they merely give an expression.
9 It is a fact that tantric pūjā, yoga, etc. are complex acts whose sense and scope are at the same time human, cosmic and divine, but which one can also consider in some cases or to a certain extent as taking place entirely on the divine plane, the gestures and operations accomplished by the officiant or adept being then only figures. (Unless it is the sacrificial act which supports the worlds – an ancient brahmanical idea that pūjā has not entirely abandoned.)
10 We find this in a number of passages of śruti (from Veda to Upaniṣads). Tara Michaël has underlined the fact in the introduction to her translation of the Haṭhayogapradīpikā (Fayard, 1974), pp. 65f.
11 We will further see how nyāsa can be used to impose a mantra which symbolises the supreme brahman: in this case one goes (or one tries to go) directly from the empirical plane to pure transcendence.
12 Tara Michaël, op cit. p. 73.
13 On the practices associated with prāṇa and nyāsa, see below. Prāṇa forms somehow a bond between the subtle and gross body of man.
14 On the nature of mantras, see the Introduction. See too Vāc, quoted above, note 6.
15 Pārameśvarasaṃhitā, Kriyākhaṇḋa (ed. U.V. Govindacarya, Śri Vilasam Press, Srirangam, 1953). The text is essentially on ritual.
Ibid. śl. 4: vyāparo mānaso hyeṣa nyāsākhyo yadyapi smṛtaḥ /
na badhnati sthitiṃ samyak tathā ‘pi kriyayā vinā //
17 See, for example, SP1, II, śl. 2–3 on the Sūryapūjā. ‘One must imagine oneself transformed into the sun’ etc. ātmānaṃ ravirūpeṇa saṃcintyā (vol. 1, pp. 72–73). These operations are described more clearly in chapter 15 of Abhinavagupta’s Tantrāloka, summarised below (pp. 60–63).
18 The level of manas and prāṇa in the sense of vital breath is that of the liṇgaśarīra, composed of three ‘envelopes’: the prāṇamaya, manomaya, and ānandamaya-kośas. Cf. below, note 35.
19 TĀ 15, 159: nyāsakrameṇa śivatāladātmyam adhiśerate. The sacrificer, the sacrifice etc. come to identify with the condition of Śiva by the process of nyāsa, that is, by the execution of the ritual procedures.
20 Or at least after a ritual activity which presupposes a certain concentration.
21 Cf. above p. 56.
22 We see here that if India does not make a clear break between body and spirit, the nyāsa is a means to integrate these two elements.
23 For example, TĀ 15, 295f. summarised below pp. 60–63. One can also refer to the third chapter of the YH, where śl. 83 indicates that the adept, who has performed the nyāsa of the śrīcakra on his body, must consider the goddess as inseparably united to himself (svābhedena vicintayet); see below p. 64).
24 There are cases, however, where meditation, evocation, or mental representation, appear distinct from placing itself. The texts can, besides, designate or not as nyāsa the operations they prescribe.
We should recall that the nyāsa, as with japa, can be considered as also having in certain cases (sometimes above all) a psychological aspect: to turn the attention towards the divinity rather than towards the actions one is performing – such is the case, for instance, for the nyāsas performed during the daily ritual bath, etc.
25 Notably by using as ink particular substances such as orpiment, sandal and so on, or by using a particular script. This, we especially meet in Chinese and Japanese Buddhism where the bījas are drawn in so-called siddha characters according to a peculiar technique which links the writing of the letters to mental concentration and control of the breath. Cf. Shuji-shu, a collection of siddham bījas, the original work of Chozen Bikshu, published by Matsumo Etasu, Koyasan, 1978. Here it is the user of the bījas who draws them.
But the adept can also use a yantra made by an other who has already written the bījas or the whole mātṛkā, so that it is enough to touch them during the ritual so as to perform the prescribed placing.
26 In his translation of these ślokas, M.N. Dutt writes that these entities ‘should be contemplated as written in fire inside the different parts of the body,’ which has no connection with the Sanskrit texts but gives a possible solution to the problem. One practitioner of nyāsa I consulted on this point told me that in the mātṛkānyāsa ‘one touches the concerned part of the body with the mahāmudrā (joining the ring finger with the thumb). While touching, one has to meditate that the letter is there in blazing form.’
27 The Tantrāloka of Abhinavagupta, with the commentary of Rajanaka Jayaratha, Kashmir Series of Texts and Studies, Bombay, 1918–1938, 12 vols. of Sanskrit text. Yoginīhṛdayam Amrtanandayogikrtadîpikâ bhâvânuvâdena ca sahitam, Varanasi, Motilal Banarsidass, 1989. This edition by Vrajavallabha Dviveda is the only reliable one of this text.
28 Abhinavagupta, in these two chapters, follows a more ancient text, the Mālinīvijayottara Tantra (MVT), a work held in special esteem by Kashmiri Śaivism, in which chapters 6 and 8 explain, briefly, the same practices intended for initiation to attain supernatural powers and salvation.
29 The samayadīkṣā is the less elevated of the initiations. After this comes the putraka(or nirvāṇa-) dīkṣā, then that of the sādhaka and finally that of the ācārya. On the subject, refer to the entries on dīkṣā of vol. 3 of the TAK. See also the article by H. Brunner ‘Le sādhaka, personnage oublié de l’Inde du sud’, Journal Asiatique, 1975.
30 Note added 2011: on this alphabetical order, see Somdev Vasudeva, ‘Synaesthetic Iconography: the Nâdiphântakrama’ in D. Goodall and A. Padoux, eds., Tantric Studies in Memory of Hélène Brunner (Pondicherry, 2007), pp. 517–550.
31 One sees here that a mantra, the sound aspect or form of the deity (and in this respect more essential than the deity) can also be considered – meditated upon – as a visible, extant, form: that of the deity itself.
The astra is the weapon of Śiva (his cult is described in the same chapter 15 of the TĀ, to śl. 378f.) – cf. infra p. 39, n. 77.
We note that a mantra can be meditated, conceived visibly in spirit, not only in the more or less anthropomorphic form of the deity or one of his attributes, but also in a written form as luminous letters, or else only as a coloured light or a luminous disc (this is especially the case in so-called Tantric Buddhism).
32 He will activate here a non-visual, phonic image which will be the subtle form of Śiva. All the gods have in principle a mūrtimantra. This is not confined to Śaivism.
33 See below p. 63.
34 Here it is the mantra itself which is the god. This is a phonic form of Śiva – but Navātman has also an anthropomorphic aspect, as a beautiful young man. It is the same for the mantra gods Bhairavasadbhāva and Ratiśekhara as, accompanied by their energy, for Mātṛsadhbhāva etc; for all deities, in fact; tantric pantheons are pantheons of mantras rather than of deities.
35 Situated precisely at the level of the energy Parāparā and forming the energy (śākta) aspect of the phonemes. See Vâc, pp. 320ff. and above, note 30.
36 Pure, says Abhinavagupta, because born from the mūrtimantra which appears (ibid., śl. 238 and 261–262) when consciousness is immersed in the pure Reality.
On the emanation within Siva, see chapter 3 of the TÂ and Vâc, pp. 306ff.
A purely spiritual and metaphysical interpretation of a nyāsa is given in the PTV, p. 20, a passage commenting śl. 27 of this text attributed to Abhinavagupta.
37 Tathā śivo’ haṃ nānyo’ smītyevaṃ bhāvayatas tathā / etad evocyate darḋhyam vimarśahṛdayāṇgamam // 270 śivaikātmyavikalpaughadvārikā nirvikalpatā /
38 This nyāsa is described in śl. 295–312 of chapter 15, which develops the information of the eighth chapter, śl. 54f. of the MVT. They are part of a section on the ‘mental sacrifice’ (manoyāga) which is to take place before the ‘external rites’.
39 The SP 1 (vol 1. p. 154f.) describes a cult of the throne of Śiva where the same series of energies and entities are invoked to make the whole cosmic manifestation the throne of Śiva. Here, the cult being mental, the adept is at the same time the seat of the divine, the divine itself, and the worshipper.
40 But laughing, conscious, his body shining like ‘ten million fires of time’ (kālāgni): MVT 8, 68, SvT 2.81.
41 Unmanā, for the Śaiva traditions, is the transcendental Void, śunyatiśunya, the supreme principle without a second (nirābhāsaṃ paraṃ tattvam anuttamam, NT 12, 23).
42 On this installation of the trident in the body for the initiation, see A. Sanderson, ‘Mandala and Agamic identity in the Trika’, in A. Padoux, ed., Mantras et diagrammes rituels dans l’hindouisme (Paris: Editions du CNRS, 1986), pp. 169–205.
43 A procedure analogous to the purification of the adhvans is described in chapters 84 and 85 of the Agni Purāṇa dealing with the nirvāṇadikṣā, chapters which are Śaivite. As described there, these rites are related to the process, in fact general in Hinduism, of the bhūta-(or deha-)śuddhi, where the elements constituting the body and cosmos are reabsorbed the one in the other up to where only pure divinity exists (for example, the LT. chapters 30 and 54: bhūtaśuddhi; Jayākhya Saṃhitā chapter 10; Kālika-Purāṇa chapters 54–55, etc.). The operation is essentially mental: it is through a meditation based on the image of the body that the elements are reabsorbed; but insofar as the process is accompanied by mantras placed on different parts of the body, nyāsas can play a role. On the process of bhūtaśuddhi, see the article by Gavin Flood, ‘The purification of the body in Tantric ritual representation’, Indo-Iranian Journal, 45, 2002, pp. 25–43.
The complex nyāsas are well represented in the texts, thus there is the mahāṣoḋhānyāsa of the fourth chapter of the Kulārṇavatantra, or the nyāsa described in the second chapter of SvT or in the eleventh chapter of the GT. The chapter 25 of the AgPur also provides an example.
The Parameśvara Saṃhitā (4, śl. 53, and chapter 54) describe a purification (śuddhi) of the body by nyāsas. On the system of the six adhvan, see Vâc, pp. 330–371.
44 One may note the frequent occurrence of the sixfold nyāsa in the Śaiva and Śākta texts. The ṣoḋhānyāsa (gaṇeśa, graha, nakṣatra, yoginī, rāśi, pīṭha) is described for example in the Vīra-tantra, the GT (chapter 10) and the Mantramahodadhi (11.48). It involves imposing six forms or aspects (rūpa) of the goddess. Cf. VMT I.1.
The TBhS 5, p. 163f. gives variants. The KT (chapter 4) has a different ṣoḋhānyāsa, the same being given by the Prapañcasārasārasaṅgraha (chapter 8).
But why the number six? The kara-, aṅga- and karāṅganyāsa which are the most current, are encountered in all traditions (see below p. 70) are also six.
45 The GT chapter 10 describes a sixfold nyāsa of the śrīvidyā composed of the six same same series of rūpa of the goddess. But this is a modern work that borrows entire passages from other texts, notably the YH. The introduction of this text published in Srinagar in 1934 is, however, interesting because it gives a table of different nyāsas with their mantra, bīja and phonemes, the points of placing and so on. This edition has run out and is difficult to find.
The Nityotsava (G.O.S. 1923, vol. 23) describes in a detailed way a laghuṣoḋhanyāsa (p. 98) which is made on the fingers, the palms and the back of the hands. It follows the complete ṣoḋhānyāsa (pp. 98–109) made on the entire body and where the placings are always preceded by dhyāna, that is, a mental visualisation of the visible aspect, described by the texts, of the imposed entities. The same text furthermore expounds (pp. 109–117), under the name mahāṣoḋhānyāsa, a placing on the body of the officiant of the śrīcakra.
On the placing of the pīṭhās see below p. 74.
46 In all this ritual, it is in effect the cakra itself that is imposed (in chapter 2, śl. 6–12, pp. 93–97, of the same text, it is the nyāsa of nine divinities presiding over nine parts of the cakra – the cakreśvarī – which is prescribed). These mantras therefore invoke parts of the cakra, and not the deities which reside there. Sometimes, however, in operations of this kind, the mantras used refer both to the cakra imposed and to the deity concerned. It also often quotes the part of the body on which a placing has been made.
47 This is only effective in the case of an adept very advanced in spiritual life: for the majority of adepts the identification remains only in the realm of representation – an observation that applies to all rites of identification described here or elsewhere.
48 There are many descriptions of the imposition of a diagram. For the śrīcakra we may quote the VM 1.112, or the Nityotsava, pp. 109–117 of the GOS edition, where the adept imposes successively the deities of the cakra on his body meditatively perceived as identical with the śricakra, as having the same shape: śarīraṃ cintayed adau nijaṃ śrīcakrarūpam. Jayaratha says similarly in his commentary on the VM, p. 60: evam tripurīkṛtavigrahaḥ sādhakaḥ.
As an instance of contemporary practices, we can quote (among many others) the ritual booklet Śrīvidyānityārcana (ed. Kalyan Mandir, Prayag, 1967). It prescribes two impositions of the cakra during the cult: first external (bahiścakranyāsa) on the body from feet to head; then internal (antaścakranyāsa) on the cakras which is purely mental.
49 In the GT, chapter 9, which is the first of the section on nyāsa of this tantra, begins with a purification of the hands (karaśuddhi) made with the mahāvidyā AṂ ĀṂ SAUḥ (śl 3–4). After this the adept purifies his fingertips with the same mantra and a red flower which he then throws away. The ṛṣyādinyāsa of the mantra (see below p. 68f.) and all the rest of the ritual are described afterwards.
In the Mantramahārṇava, a vast collection of Śaiva and Vaiṣṇava mantras which describes worship and practice, the ṛṣyādinyāsa – which places as it were the preliminary mantras on the body of the celebrant – comes first, but it is immediately followed by the karāṅganyāsa preceding the worship itself.
50 That is, makes full or complete (sakala), or which provides the body of the officiant with parts (kalā). This is another name for the karāṅganyāsa. This can be found in a Sanskrit text quoted by Aghoraśiva, with a French translation, in appendix 1 (pp. 322–325) of the first volume of the edition of the SP by Hélène Brunner.
In a general manner, the rite of the sakalīkaraṇa consists of imposing the aṅgas of the deity on the body of the disciple to divinise it and to create, in the course of the pūjā, the body with its diverse parts (sāṅgam) as the deity to be worshipped.
51 On the different meanings of this term, and especially on the notion of portions or parts of a deity, see the long entry on kalâ of the TAK, vol. 2.
52 Twelve is, in fact, a solar figure. It is the number of kalās of the sun, whereas sixteen is the number of kalās or digits of the moon.
53 Pārameśvara-saṃhitā 4.5–22 (p. 18): I have translated the passage, simplifying it a little. The JayS 11.10–22 (pp. 94–95) prescribes practically the same nyāsa: from the mūla and mūrtimantra, then the aṅgamantras hṛdaya etc. and the attributes and aspects of Viṣṇu. The three final śloka are analogous to those which end the passage translated here. The LT 25.61–63 (pp. 135–136) prescribes nyāsa first on the fingers with five śakti, then with the aṅgamantras, and following with the placing of the aṅgas of Viṣṇu on the palms. See also Agastyasaṃhita 2, chapter 12 etc. In the Śaiva domain, one can cite the Tantrasāra of Kr̥ṣṇānanda, p. 74 etc, A karaśuddhi made with the mūlavidyā of the Goddess is also prescribed before the pūjā in the VMT 1.109 (p. 57). A mūrtinyāsa, distinct from the karanyāsa, is prescribed in the JayS. 20.180–181 (p. 212).
The placings can be of deities, of aspects of a god or, in the case of a mantra, of different words of the mantra etc. One can also place serially groups of phonemes of the mātr̥kā, this is one of the forms of the mātṛkānyāsa. The procedures differ according to the schools, traditions etc; it does not seem, at first sight, that they can be reduced to a general rule. There is even in the Mantramahārṇava 1.6 (folio 95f.) a mantra of Tryambaka of a vedic type: Oṃ tryambakam yajāmahe sugandhiṃ puṣṭivardanam urvarukamiva bandhanānmṛtyor kukṣīya māmṛtāt, of which the karanyāsa is made in the following way: oṃ tryambakam aśṅguṣṭhabhyāṃ namaḥ, yajamahe tarjanibhyāṃ namaḥ, sugandhiṃ, puṣṭivardhanaṃ madhyamabhyāṃ namaḥ etc. up to: māmṛtāt karatalabhyāṃ namaḥ.
54 TĀ 15.456–459 (vol. 9 pp. 230–231).
55 SP2, pp. 52–53. The text is brief, a single śloka, but Helene Brunner explains it in a note, referring to the Purusaṃhitā itself cited in the commentary of the Mrg, kriyāpāda 7.14, which mentions the rite. One washes the right hand, anoints it, imposes the brahmamantra on it, then one draws, with the thumb of the left hand on the palm of the right, a maṇḋala where one places Śiva, to which one pays homage. See also SP3, p. 289. or the Suprabhedāgama, chapter 4 analysed by Brunner (Journal Asiatique, 1967).
The Buddhist hastapūjā (Finot, ‘Manuscrits sanskrits de Sādhana,’ Journal Asiatique, 1934), includes the nyāsa of bījas, mentally placing them on the hand together with a lotus and divinities; this is a rite akin but different. The hastābhimantraṇa mentioned in the Adikarmapradīpika, edited by L. de La Vallée Poussin (Études et Matériaux, pp. 192 and 217) appears, on the contrary, as a hastanyāsa, although the word nyāsa is not used: hastam abhimantrya [mantraṃ] vibhāvya.
56 In the sense, it is explained, of being opposed (vāma) to transmigration (śl. 278) and of secrecy (śl. 279) as is normally the case of the practices of the so-called left hand, in particular in the Kaula schools.
57 A gesture one sees often made by Indians outside of a ritual context. Generally, thepositions or movements of the hands of Indians differ from those observable in Western societies.
58 According to the commentary of the SP1 – III.41 (vol. 1, pp. 142–143 note), the āsanamantra involves putting a flower on the throne which is thus consecrated. This resembles the Nityotsava (nyāsaprakaranam, p. 95) where a triple āsana must be imposed with flowers on the śricakra.
The GT as cited in Principles of Tantra, p. 158f., seems to have a rite of installing a devatā in a yantra by means of flowers to which the adept has transmitted divine energy with his breath and the vāyubīja YAṂ.
59 See, for example, Dale Saunders Mudrā (New York: Pantheon Books, 1960), or Stephen Beyer The Cult of Tārā (Los Angeles: University of California Press, 1973).
60 The SP, for example, prescribes a number of mudrās, and in the first volume of Hélène Brunner’s edition there is an appendix of descriptions and photographs. One can also cite āgamas and saṃhitās etc. In volume V of the History of Dharmaśātra (pp. 1123–1131), P.V. Kane presents a brief study with an abundant – but still far from complete – Sanskrit bibliography. For Buddhism, one must add to India, Tibet, China and Japan.
61 I have not checked their provenance. There are several texts, from the non-dualist Śaiva Krama tradition, of that name.
62 As the royal seal is a symbol of authority, the mudrā symbolises the power of the divinity; or as the seal authenticates a contract, the mudrā seals a pact between the worshipper and the worshipped. Cf. Dale Saunders, ‘Symbolic Gestures in Buddhism’ Artibus Asiae vol. 21, 1957.
63 In the context of Buddhism, a justification of the mantra-mudrā association has sometimes been based on the theory of the three bodies, vāk-kaya-citta: these being three coordinated and complementary aspects of an ensemble.
64 Bhāvayati, causative of BHU, means to make become, give birth, but also to consecrate, to control and to have in mind, to consider, and to know. The word bhāvanā (f.), derived from BHU, designates intense, mental concentration, the creative power and efficacy of imagination, and mystical realisation (in the ‘way of energy’, the śāktopāya according to the TA). On bhāvanā, François Chenet, ‘Bhāvanā et créativité de la conscience’ (Numen, vol. xxxiv, 1987, pp. 45–96).
65 In this I entirely disagree with Frits Staal’s position: see his Ritual and Mantras. Rules without Meaning (Delhi: Motilal Banarsidass, 1996).
66 We know that the word mudrā has several senses, one of which is that of a mystical attitude; mudrā is thus the seal of the ineffable, purely spiritual; in this sense see Mahārthamañjarī (MM), cited in the next note; see also TĀ 32.1–3 – on which A. Padoux, ‘The body in tantric ritual: the case of the mudrās’, in Teun Goudriaan, ed., The Sanskrit Tradition and Tantrism (Leiden: Brill, 1990).
67 Avikalpatayā marś vikalpavargasyāṅgasannāhaḥ / Arghyaṃ vedyavilāsaḥ puṣpāṇi svabhāvapoṣakā bhāvāḥ. The reference is: Maheśvarānandapraṇitā mahārthamañjarī svopajñānaparimalākhyavyākhyopetā (Varanasi: Vārāṇaseya, Sanskrit Viśvavidyālaya, 1972), pp. 112–113.
68 These elements are not always mentioned. The ṛṣi, the chandas, and the devatā are indispensable and their mention is ancient (along with the viniyoga), dating back to the anukramaṇi of the Veda. The other elements are Tantric and they do not always appear; they cannot therefore be considered as necessary elements of all Tantric mantras.
The ṛṣi is, of course, entirely fictitious; it is in fact sometimes a deity. As for the chandas, it can be real if the mantra is of Vedic origin or if the Tantric mantra’s structure corresponds to a certain meter. But the non-Vedic mantras are often formed from a group of short syllables (often unpronounceable) or even consist in one bīja, such as HRĪṂ or SAUḤ, or the fifty varṇas of the mātṛkā, which do not, strictly speaking, have a meter. There are, however, lists of ṛṣi and chandas for each of the fifty varṇas. In other cases, several meters are given. This mention is therefore preserved in many cases only as a kind of reference to the Vedic tradition. The reference is real when elements of a Tantric mantra are derived from a Vedic text, in which case these will be given the same meter as the Vedic passages from where they come. Otherwise, the reference seems to be purely theoretical; although, for contemporary followers, the chandas appears to expresses, if not really a meter, at least a number, an inner rhythm particular to the mantra: for such adepts, this is nothing theoretical.
It sometimes happens that the bīja is formed from one or from several words, which is contradictory. The śakti is, in theory, the central part of the mantra and is believed to concentrate in itself its energy; its place in the mantra is in fact variable. As for the kīlaka, a pin, bolt or wedge, it should normally be at the end of the mantra as in an arrow so as to pierce its target, but its place and form also varies.
Another element, the kavaca (in the sense of amulet or protective formula) is sometimes added.
69 This is the principle, valid even if this transformation is generally only theoretical and imagined.
70 We see here the usual dialectic of the tantric pūjā. deva eva yajed devam, nādevo devam arcayet, that one finds constantly and which also explains the purifications during the pūjā as well as the succession of deifying nyāsas (those which we have previously seen for example). One must be already deified to be permitted to ritually worship a deity by a cult which is deemed to lead to deification, and one must be already ‘mantricised’ in order to use and assimilate more completely a mantra. It is necessary to have, in some measure, at the beginning that towards which one strives.
Should we not see there a sign of what is, it seems to me, a well-pronounced intellectual tendency in India: a tendency not to admit (or a difficulty to conceive) a first beginning or a progress which would be a new advent: before starting, things are as it were already there. This is the view, so widespread, of the satkāryavāda, for which the effect pre-exists in the cause or, for cosmology, of all cyclical cosmologies.
Of course, insofar as the pūjā continues – and preserves certain elements of – the old, brahmanical sacrifice, the preliminary divinisation of the tantric officiant can be regarded as continuing, under a different form, the rites by which the Vedic sacrificer and the victim were preliminarily consecrated in order to enter into communication with the gods in the sacrificial area. Perhaps we should add that, in the course of the ritual, the nyāsas seem to have only a temporary role and effectiveness: in each part of the rite, placings are done that are cancelled by those which follow, or by the ācamana which separates them (an observation that I owe to Hélène Brunner). The cancellation of a rite by the following one(s) is also one of the reasons of the redundancy so characteristic of Hindu rites.
71 This according to the Mantramahārṇava 1.6 (folio 91).
Some texts specify the mudrā to be used for these placings. Thus in the Śārasaṅgraha cited in the TBhS, 5, p. 70: the placings of the ṛṣi, chandas and devatā are done with four fingers excluding the thumb – the same way for the bīja and the śakti. But the kīlaka, according to another author, is imposed with both hands on all the body: it is the same if, as is generally the case, the viniyoga finishes the series.
72 For a better view of the mūdras shown during a pūjā, see for instance G. Bühnemann, The Worship of Mahāganapati according to the Nityotsava (Witrach: Institut für Indologie, 1988); or Richard H. Davis, Ritual in an Oscillating Universe. Worshipping Siva in Medieval India, Princeton: Princeton University Press, 1991).
A small modern treatise, the Yoga-mantra-saṃhitā, explains: the ṛṣi who has ‘seen’ the mantra is for the adept as his master: that mantra is placed on the head with respect; the meter being made of words and of syllables is placed on the mouth; the deity, residing in the heart of all creatures, must be placed in the heart; and the seed on the genitals.
73 See the translation by Brunner (SP vol. 1, p. 132). Here is the text:
Hṛdayādikarānteṣu kaniṣṭhādyangulīṣu ca / hṛdādimantravinyāsaḥ sakalīkaraṇaṃ matam //
This nyāsa is also called ṣaḋaṅganyāsa, hṛdayādinyāsa, or hṛdayādiṣaḋaṅganyāsa and, for the hands, anguṣṭhādiṣaḋaṅganyāsa.
The adept must make the aṅgamantras according to the rules at the same time as he places them, thanks to which he deifies his hands and body. Sometimes netra does not appear among the aṅga, which are then only five: this is the pañcāṅganyāsa prescribed in various texts – thus the Nāradapañcaratra 3.3,10–11.
74 There are also secondary aspects called upāṅga: accessory limbs which are placed after the first ones.
75 They are also sometimes represented as separate deities that one ritually places on or around the image being used for worship – thus SP1 (vol. 1 pp. 208–210). On the aṅga and their symbolism in Śaivism see SP1, introduction, p. xxxiv and pp. 194–197), also vol. 3, pp. 400–404 (note 440), and Hélène Brunner, ‘Les membres de Śiva’, Asiatische Studien/Etudes Asiatiques, XL.2 (1986), pp. 89–132.
76 In the Śaiva tradition, this is the third eye of Śiva, that of consciousness, situated on the forehead between the two others eyes: the word netra in the aṅgamantra netra is in the singular (for the third eye) or in the plural (for the three eyes), but never in the dual.
77 The consonants which serve to form these bījas vary according to the deity, the school etc. But the vocalic element is stereotyped – saḋdhīrghair ādyādyabījena karāṅganyāsam says the Phetkāriṇitantra, as cited in the TBhS (p. 170). One forms the bījas on the basis of the bīja or of the principal vidyā (ādyabījena) of the mantra while associating it, in due order, with the six long vowels Ā etc. (ṣaḋdīrghair ādi…), in fact with ā, ī, ū, ai, au, and aḥ. One will thus have, starting from the vidyā SAUḤ: sāṃ, sīṃ, sūṃ, saim, saḥ; starting from HṚĪṂ: hrām etc.
The vaktra mantras, on the other hand, are done with short vowels. The same instructions are in the Agni Purāṇa 25, 5 fl. (Vaiṣṇava chapter); the same (for only the aṅgamantra) in the JayS 6.105ff. etc.
78 In the Pāñcarātra tradition, the LT (ch. 33) gives the aṅgas of the tārikāvidyā of Tārā used in the cult of Lakṣmī: one adds to the aṅga a term (jñāna, aiśvarya, śakti, bala, tejas) whose invocation carries to the adept, through nyāsa, each of the qualities enumerated – the aṅgas are of the model Oṃ hrāṃ jñānāya hṛdayāya namaḥ (we note that the aṅgamantra for netra is netrabhyām, the two eyes, instead of the three eyes of Śaivism). The commentary on the Śaiva ST 18.6 (p. 713) associates in the same way to the six aṅgas the enunciation of the six qualities (guṇa) or infinite glories of Śiva, omniscience, etc. The SP1 quoted above associates the aṅgamantras with five other attributes of Śiva (the netra is sometimes omitted – cf. above note 68.
79 Phaṭ is usually called the astra (or heti) mantra – it is the weapon, the missile, especially the arm of Agni. It shines, says the TĀ (15.232), as the flame of the fire of time, destroying and purifying the gross and subtle body of the adept. The same designation of phaṭ occurs in Buddhism – cf. Finot, ‘Manuscrits sanskrits de sādhanas’, JAs. 1934, p. 60: phaṭ phaṭ sarvāṅgeṣvastram … arpayet.
80 The rule often allows exceptions in certain traditions. See, for example, the notes of N.R. Bhatt in his edition of the Raurāvāgama (I.1.4, vol. 1, p. 18); see also the commentary of the Haṃsopaniṣad by Upaniṣadbrahmayogin (ed. Adyar, p. 566). There may also exist variants in the case of optional rites (kāmyakarmāṇi).
81 Preceded sometimes by a bīja formed as indicated above.
82 This is according to the TBhS (p. 170). But there are other ways of doing it: thus the Nāradapañcarātra III.3.24. This triple nyāsa has the main sense of symbolising the threefold activity of the divinity and therefore of placing it with all its cosmic dynamism. But each of the three nyāsas can also be supposed to have a particular effect (cf. Nāradapañcarātra loc. cit.) or else can be prescribed for different categories of adepts (cf. Tāriṇīpārjāta, p. 23).
83 One finds the aṅga-or karāṅganyāsa described or prescribed as well in texts other than those presented above or in the preceding notes. For example, the TĀ (chapter 30), the SvT 1.71–72, the NT 12.28–33, the Kālīkāpurāṇa, the JayS 6.105ff. etc.
Mantrā mūkatvam āyānti vinyāsena vinā lipeḥ/
Sarvamantrasiddhyarthaṃ tasmād adau lipiṃ nyaset// (TBhS p. 159).
A small modern manual (Śrīvidyānityāḥnika) describes the mātṛkā as a great mantra serving to place the goddess Mātṛkāsarasvatī: śrīmātṛikāsarasvatīnyāsamahāmantra.
85 ‘The mātṛkā, mother of the gods, origin of all mantras’, says the GT 9.14, in the exposition of nyāsa. And Abhinavagupta in the TĀ 15.130–31: ‘This inner energy born from Bhairava under the aspect of the mass of sound (śabdarāśi), one calls mātṛkā, because it is the mother of all that appears’ (TĀ vol. 9, p. 58).
86 The formula used is generally of the type: a group of consonants (varga) framed by two vowels (one short, one long), these and the others with the bindu: aṃ, kaṃ etc. + part of the hand or aṅga + the appropriate jāti (namaḥ etc.); this allows the placing of twelve vowels (6 × 2) and all the other phonemes. Cf, for example, Nityotsava (nyāsaprakaraṇam), p. 73.
87 Vaiṣṇava Gautamīya cited in TBhS p. 160. The placings on the lotuses are not necessarily made in the order from bottom to top, as in the Dakṣinamūrti-saṃhitā cited by TBhS p. 160. This saṃhitā, which is a text of the Śrīvidyā and not of the Pāñcarātra, contains in its eighth paṭala the description of a mātṛikāpūjā including ṣaḋaṅga- and bahirmātṛkānyāsa, thanks to which, it tells us (śl. 16), the mantrin will become varṇa-svarūpa, thus identified in his essence with the divine energy of the phonemes. This pūjā continues with the dhyāna of the mātṛkādevī, then by the drawing of a yantra where worship will be performed to her.
88 TBhS p. 160. The same in the Nityotsava p. 94: iti dhyatvā manasā puṣpāñjaliṃ dattvā mātr̥kāḥ … svāṅgeṣu nyaset.
89 TBhS citing the Vaiṣṇava-Gautamīya. This description, as the name of the text would suggest, is a Vaiṣṇava work. I have not, however, found in the Vaiṣṇava texts I have consulted a description of the mātṛkānyāsa being exactly the same as that of the Śaiva texts I quote here, which is more complex.
90 Citations of another tantra (tantrāntara) in the TBhS 5, p. 159.
91 The prescriptions are varied. But often the vowels are placed on the face and head, the consonants on the body, and the semi-vowels, spirants and KṢA on the heart, associating them there with the elements that constitute the body, the breath, and the self. Thus the GT 9.25–31; AgPur, chapter 293, 37–39; lipinyāsa; Tāriṇīpārijāta, pp. 18–19 etc.
92 TBhS, pp. 160–162.
93 Cf Vāc pp. 320–327, based on TĀ 15.120–142 – vol. 9, pp. 62–74. The mālinī is also mentioned in other, earlier tantric works, such as the Siddhayogesvarimata, the Tantrasadbhâva, the Satsāhasrasamhitā, etc.
On the subject I refer to Somdev Vasudeva’s interesting and novel study, ‘Synaesthetic Iconography: 1. the Nâdiphântakrama’, in D. Goodall and A. Padoux, eds., Mélanges tantriques à la mémoire d’Hélène Brunner, Pondichéry: Institut Français de Pondichéry et Ecole Française d’Extrême-Orient, 2007, pp. 517–550. Note that the above paragraph on mālinīnyāsa was re-drafted in 2009.
94 Cf. the study of D.C. Sirkar, The Śākta Pīṭhas (JRASB letters, XIV, 1948, pp. 1–108 and MLBD). It is a presentation and edition of a text, the Pīṭhanirṇaya, or Mahāpīṭhanirūpaṇa.
A special number of the Hindi periodical Kalyān dedicated to pilgrimage (‘Tīrtāṅk’, Gorakhpur n.d.) gives (pp. 513–528) a list of 108 places of the cult of Śakti, then of the fifty-one pīṭha of the goddess, with a map and descriptions of the places, followed by a study of the secret pīṭha of the goddess: śaktipīṭharahasya.
The TĀ (15.81–98) enumerates the pīṭha ‘external’ or ‘internal’, that is, situated in the world or interiorised in different points of the physical or ‘subtle’ body.
95 The system of 108 pīṭha has in effect been taken over and adapted in the Vaiṣṇava tradition.
We may note, apropos of pīṭhas, that some texts (thus the ŚT 2.134 comm; vol. 1, p. 115) call the sixteen vowels pīṭha, as contrasted with the consonants then called kṣetra. We see here the connection between phonemes and space. Jayaratha, in fact, commenting on the first śloka of the VM where the goddess is described as assuming six aspects (Ganeśa, rāśi etc.) which are placed in the ṣoḋhanyāsa, stresses the point that the pīṭhas correspond to extension since the goddess also appears in this form (deśrūpatayā): ‘The Lord,’ says Abhinavagupta in the ĪPV 2.1,5, ‘manifests spatial succession by the diversity of forms (mūrtivaicitrya)’; this corresponds to the path of space (deśādhvan) in the theory of the six ways (TĀ 6.34–36).
The diṅnyāsa, the placing of the directions of space, which is a part of several rites, does not seem to be associated with the same notion since it consists in reciting the formulas of the nyāsa and showing a mudrā while turning to the eight directions of space in order to win the favour of the devatās of the directions; this is not really a nyāsa.
96 We also find a brief description in the Mantramahodadhi which concludes: iti dehamaye pīṭhe dhyayet svābhīṣṭa-devatām: ‘he should meditate his own deity on the seat which has become his body’. Also see Tāriṇipārijāta pp. 34–37, which enumerates the entities and mantras to be placed; Tārarahasya 3, p. 119; Siddhasiddhāntapaddhati 2.1–9; or a recent manual Balārcanapaddhati and so on.
97 That is to say, from the twenty-sixth to the thirty-fourth tattva in the system of thirty-six tattvas of Tantric Hinduism.
98 It is perhaps interesting, from the point of view of the range or psychology of the nyāsa, to mention here a poem cited in the Principles of Tantra (p. 1142), where the sixth verse, recalling the unity of the worshipper with the divinity he worships, adds: ‘otherwise, how does Nyāsa on Thy limbs affect Me or Nyāsa on my limbs affect Thee?’ ‘They are one and the same,’ comments A. Avalon, ‘and therefore when the sādhaka does Nyāsa on himself, he does Nyāsa on the Devatā also’ (ibid. note 4).
Another point of view, contemporary but undoubtedly exactly traditional in sentiment, is that of R.K. Ray Encyclopedia of Yoga (Varanasi, 1975) which, in relation to pīṭhanyāsa, writes: ‘Nyāsa is useful in effecting the proper distribution of the Śaktis of the human frame in their proper positions so as to avoid the production of discord and distractions in worship. The actions of Nyāsa on the body, finger and palms stimulate the nerve centres and nerves therein. Thus, for the attainment of that state in which a Sādhaka feels that the Bhāva (nature, disposition) of the Devatā has come upon him, Nyāsa is a great auxiliary. It is, as it were, the wearing of jewels on different parts of the body. The Bījas of the Devatā are the jewels which the Sādhaka places on different parts of his body. By Nyāsa he places his Abhiṣṭadevatā (Iṣṭa-devatā) in such parts, and by Vyāpaka-nyāsa (comprehensive Nyāsa) he spreads His presence through himself and he becomes permeated by Him, losing himself in the divine Self.’ (See too MNT 5.105 and passim; ŚT 6.1ff; Prapañcasāra Tantra p. 138 and passim; Yoginī Tantra 2.1.1 and passim).
99 The nyāsa can be made with a flower or blades of grass as said before – see p. 71.
100 SP1 (vol. 1, pp. 124–125): darśayitvāmṛtīmudrām ‘and one shows the mudrā which transforms into nectar’ (Brunner examines this problem in the introduction, p. xxxvi).
101 The Tāriṇīpārijāta, pp. 18–19, describes, after expounding the bhūtaśuddhi, a jīvanyāsa made with the prāṇapratiṣṭhāmantra of the goddess Tāriṇī along with a whole series of placings (ṛṣyādi, kara, ṣaḋaṅga etc.) with dhyāna, japa etc. The structure of the rites is the same as that of the prāṇapratiṣṭhā such as those found in diverse modern manuals (Śrīkapadruma or some other).
The jīvanyāsa is mentioned in the same way in the MNT (5.105). The adept places by this rite the divine breaths in his own body: taddehe devyāḥ prāṇān nidhāpayet (the commentary glosses taddehe by tasmin navīne dehe: ‘in this new body’, in this body renewed, transformed by the bhūtaśuddhi.
The same occurs in SP 1 (vol. 1 pp. 128–133).
102 The vyāpakanyāsa consists usually in passing the two hands several times on the entire body, from feet to head, while repeating the mūlamantra (see TBhS 5, p. 171). We have seen that this can also be done differently. But it always consists in the placing of a mūlamantra, or of a fundamental entity or diagram, on the entire body, which must be entirely impregnated and penetrated (vyāpta).
Cf. Principles of Tantra p. 1140: ‘and then by means of Vyāpaka, or comprehensive Nyāsa, to feel the presence of the devatā as one undivided entity whose substance is Mantra all over one’s body from the feet to the crown of the head’.
103 See, for example, the LT chapter 46: pūjā of the laksmīmantra. Three other chapters of this text (47, 48, 49) describe the cult of other mantras: the nyāsas are identical. For the tārikā-mantra, the principal mantra of the Tantra (chapters 42–43), different nyāsas are prescribed, but this is because there is no ‘external’ worship here but a sādhana where mental elements, visualisations, japa, and mantrayoga, play an essential role.
Other Vaiṣṇava examples include: the pūjā of Vasudeva-mantra with eleven akṣara in the Varśakriyākaumudī pp. 182f.; or the placings of the mūlamantra of eight and twelve akṣara in the Sanatkumārasaṃhitā I, adhyaya 9 (p. 72f. ed. Adyar, 1969).
104 R.V. Joshi Le rituel de la dévotion kr̥ṣṇaïte. Pondichéry: Institut Français d’Indologie, 1959 (publication of the IFI, no 17).
105 A short text published, translated and annotated by F. Nowotny in the Indo-Iranian Journal I. 2, 1957, pp. 109–154.
106 Śrībalārcanapaddhatih … śrī paśupateśvarānandanāthaviracitā, Surat, sam. 1991 (small text of ninety folios).
1 The French original version of this paper, ‘Un rituel hindou du rosaire (Jayākhyasaṃhitā, chapitre 14)’ was published in the Journal Asiatique, CCLXXV, 1987, pp. 115–129.
2 Tantric texts are normally meant for the instruction of initiates, these being generally sādhakas, adepts looking for rewards or powers (bubhukṣu). On the sādhaka, H. Brunner’s study, ‘Le sādhaka, personnage oublié du śivaïsme du sud’ (J.As, 1975, pp. 411–443), is still very much worth consulting. The Śaiva sādhaka- or nirvāṇa-dīkṣ ā is described in vol. 3 of H. Brunner’s edition of the Somaśambhupaddhati – SP (Pondichéry, 1977).
3 One may of course acquire a fully constituted rosary: this is what all non-initiates – the majority of devotees – do. But such a rosary must be consecrated before it is used.
4 Perhaps surprisingly in a Vaiṣṇava context, the rudrākṣa, the berries of Śiva’s tree, are considered the best ones.
5 Tantras usually count women and children among those seeking mokṣa.
6 Astra, the Weapon mantra, is used in many ritual contexts, being placed (nyāsta) or ‘sprinkled’ (saṃprokṣya).
7 saṃpūjya puṣpadhūpādyais tasya śuddhim athā caret //28// dagdham astreṇa saṃcintya varmaṇā māruteritam / āplavyamūlamantreṇa paramāmr̥tarūpiṇā //29//. Astra and Varman are two of the six ‘limbs’ (aṅga) of Viṣṇu (or Śiva). Astra, the Weapon, is a divine destructive arrow. Varman, or Kavaca, the Cuirass, is protecting: it safeguards the purity given by the fire of the Astra deemed to burn away all impurities. Bathing and fanning are two of the usual services (upacāra) offered to a deity. We may note here the important role of mental representations in the Hindu cult, the pūjā, which is to a large extent a play of imagination.
8 Caturbhujaṃ tu virajo nārāyanam ivāparam //30// varadābhayahastaṃ ca baddhāñjalidharaṃ smaret/brahmadvārasthitaṃ tac ca sūtraṃ dhyayecchikhopamam //31//. The mentally conceived divine anthropomorphic figure of the rosary is imagined as bearing it also on its head as is the case of the āyudhapuruṣas who represent symbolically the ‘arms’ of a deity and bear them on their heads.
9 On the ritual for the Tantric worship of a deity, see SP, vol. 1, pp. 90ff.
10 On the uddhāra of mantras, see the first paper, above,
11 In the introduction, p. 37, the editor of the JS gives a shorter form of the mantra: OṂ THAṂ aksasutrāya namaḥ.
12 On the aṅgas, see H. Brunner’s study: ‘Les membres de Śiva’ in Asiatische Studien/Etudes Asiatiques, XL. 2. 1986, pp. 89–132.
13 On bhāvanā, see F. Chenet, ‘Bhāvanā et créativité de la conscience’, in Numen, vol. XXXIV, 1987, pp. 45–96.
14 sr̥ṣṭikrameṇa, since the power of the deity is entering the image of the rosary, a process analogous to that of the cosmic process of emanation (sr̥ṣṭi).
15 Yā parā vaiṣṇavī śaktir abhinnā paramātmanaḥ //35// pracalatpūrṇacandrābhā sūryavatkiraṇāvr̥ttā / yugakṣayograhutabhuktejasā tīvrabr̥ṃhitā //36// prāgvat sr̥ṣṭikrameṇaiva smared hr̥tpadmamadhyagām / pūrakeṇa tu viprendra kumbhakena nirodhitām //37// hr̥tpadmadmād utthitāṃ bhūyo brahmarandhrāvadīṃ smaret/brahmarandhrāt tato vipra prollasantīṃśanaiś śanaiḥ //38// pūrvavat sandhimārgena cintayed dvādaśāntagāṃ / recakākhyena yogena tayā sūtraṃ tu bhāvayet //39//
16 Pr̥ṣṭhe: that is, by the part opposite to the Meru.
17 Note that the JS teachings are only meant for members of the three higher varnas.
18 niṣkalo mantrarāt tataḥ //40// mantravr̥ndasamāyukto magnas tatreti cintayet/ mantrātmā bhagavacchaktis sūtramantrākṣarāṇi vai //41// sarvam ekīkr̥tam. dhyāyed yathā kṣīreṇa sodakam / tatra puṣpāñjalau pr̥ṣhe gr̥hītvā cintayed imam //42// susaṃ sr̥tākṣasūtrasya patantaṃ pr̥ṣṭhato dvija / sr̥sṭikramāt samāyātaṃ mantraṃ sakalaniṣ kalam //43// sphuliṅgaṇasaṅkāśam ādheyatamutāṃ gatam /
19 idaṃ vijñāpya deveśam aksasūtramayācyuta //44// prayaccha mantrajāpārthaṃ smared dattaṃca tena tat / gr̥hītvā śirasā paścāt prasīda iti coccāret //45//.
20 prāptaṃ saṃsrāpayet paścād arghyapātrāc ca vāriṇā / saṃpūjya puṣpadhūpādyair mantraṃ tac ca vinyaset //46// sādhāraṃ sādhanaṃ caiva śaktipūrvais samāvr̥ttam / sannidhau bhava deveśa sanniruddho bhavācyuta //47// sūtrākhye maṇijāle’smin yāvac candrārkatārakam /.
21 The four stages from thought to action thus correspond to the planes of consciousness, from turya to jāgrat. The inverse movement is that of the return from discursive waking state to pure inner consciousness.
22 smaraṇa is not mere remembrance or memory. It is a recalling to mind and a mental concentration on a thought or on a mantra. In the Trika, smaraṇa is used to designate the intuitive consciousness of the highest Reality. In the PTV (p. 245), Abhinavagupta glosses smarati with anusaṃdhatte, that is, one-pointed attention. This sense given to smarati goes back to the ancient meaning of SMR̥: on this, see L. Silburn, Instant et Cause (Paris: Vrin, 1955, pp. 374–346).
23 śl. 52b–64: karaṇaṃ vāṅmanaścaiva mantreṇa paramātmanā //52// viṣṇunāśaktirūpeṇa bhāvitaṃ bhāvayet purā / cetasā yad uparūḋhaṃ tad yuktaṃ vastu vāggatam //53// vastuyuktaṃ ca caitanyaṃ vāktr̥tīyā ca nārada / samārohet karmapade vikāratve śanaiś śanaiḥ //54// evaṃ vivartate mantras turyāj jāgrāvadhi kramāt / prātilomyena vai vipra punar eva nivartate //55// karaṇaṃ vāksvarūpaṃ syād vāk ca cidrūpiṇī bhavet / yac cittaṃ sa bhaven mantro yo mantras sa tvajo hariḥ //56// brahmādya jāgratparyantaṃ punar atraiva samnayet / vyāpakaṃ yat paraṃ brahma śaktir nārāyaṇī ca yā //57// sā hy eva pariṇāmena turyākhyaṃ bhajate padam / turyaṃ suṣuptatām eti suṣuptaṃ svapnatāṃ vrajet //58// jāgratvaṃ svapnam āyāti evaṃ jāgrādaditaḥ punaḥ / bhagavacchaktiparyantaṃ sandhānaṃ caikataṃ smaret //59// yad anityam idaṃ vipra citravat parid śyate / bāhyaṃ viṣayajālaṃ jāgrad etad udāhr̥tam //60// anityapratipattir yā asminn upari sarvadā / bhaṇgure svapnatulyo yas sa svapno jāgrakāraṇam // 61//suṣuptaṃ śāntatāṃ viddhi svapnavr̥tteḥ paraṃ tu yat / śaktirūpasya vai viṣṇoḥ prāptas turyatvam eti saḥ //62// tayā saha samatvaṃ ca turyātītaṃ tad ucyate / mantrātītaṃ paraṃ mantraṃ sthūlasūkṣmadvayaṃ tathā //63//ātmānaṃ pañcamaṃ vipra ekatvenānusaṃdhayet / kr̥tvaivaṃ anusaṃdhānaṃ prārabheta japaṃ tathā //64//
24 hr̥tpuṇḋarīkamadhyastho bhārūpaḥ parameśvaraḥ / nirmalasphaṭikaprakhyaḥ prasphuran yaḥ svatejasā //65// tasya śabdamayī śaktī (? ?) jvālāvan nismr̥taṃ mahat(?)/taddharmadharmiṇī śuddhā tasyā vai varṇasaṃtatiḥ //66// nismr̥tā mantrajananī tadagrāc- caiva mantrarāṭ/nismr̥taṃ (tu) yathā puṣpaṃ latāgrān munisattama //67// evam evākṣasūtre tu uditaṃ ca kramaṃ smaret/hr̥llayam. tallayīkuryāddhr̥llayasya ca tat thatā //68// ekasya mantranāthasya antarbāhyoditasya ca /.
25 Such a rite for the so-called ‘outer worship’ (bāhyayāga) of the deity is described in chapter 13 (śl. 100ff.) of the JS. A similar ritual is to be found in the Śaiva context – (see for instance the SP1 – similitude due to the fact that the Pāñcarātra was strongly influenced by shaivism.
26 This is carefully described in śl. 73–76a. It appears to be done by a particular movement of the thumb which moves the rosary along the finger (the index, normally) on which it rests, the counting going on in the same sense. To avoid crossing the Meru, other texts prescribe counting the beads alternatively in one sense and then in the reverse.
27 There are variant lists of these acts, some including nine acts. On the ṣaṭkarmāṇI, see T. Goudriaan’s study, Māyā Divine and Human, already quoted, pp. 251–412.
28 The passage I allude to is as follows: « paradehepraveśe ca tattvānāṃpreraṇe dvija/ krūrakarmaṇī siddhyarthaṃ japo jyotirmayas smr̥taḥ // 79// tatrāntarlīnam amalaṃ śabdaṃ tu paribhāvayet / bhuktimuktiprasiddhyarthaṃ doṣaduhkhakṣayaṅkaram //80// śāntau tu sarvakāryāṇāṃ siddhyartham avicārataḥ / sukhasaubhāgyasiddhyarthaṃ tataḥ ‘py ādhānakarmaṇi //81// japo bhavati śabdākhyas tatrāntaras tu vibhāvayet / niśāmbukaṇasaṅkāśaprakāśātma janārdana //82// hr̥tpadmāt tu svamantreṇa bhāvayeccchabdam utthitam / nityoditaṃtu tenaiva prayatnarahiteṇa tu //83// praṇāpannapadasthena bhāvanāmiṣritena / japyamānas tu vai mantras sarvakāmaphala pradaḥ //84//.
29 In all ritual worship, the deity first installed in the icon by the rite of āvahana so as to be worshipped is removed from the ritual icon by the visarjana rite, once the worship is over. The same process is used for the rosary. The mantra being used is: OṂ bhagavanmantramūrte svapadam āsādya kṣamasva kṣamasva OṂ namo namaḥ.
30 This mantra is: OṂ ḤRŪṂ VAUṂ viṣvakṣenāya namaḥ.
31 sandhānam upasaṃhr̥tya pūrvam. yac cākṣasūtrakam / visarjanākhyamantreṇa mudrāyuktena nārada //90// sūtraṃ cābhinavaṃ kr̥tva prāgvat saṃskr̥tya veṣṭayet / yathāsthitaṃ māṇināṃ ca yojanaṃ cācaret punaḥ //91// saṃsthāpya vidhivan mantrī bhūyas tasmiṃs tu vinyaset/purā yad āhr̥taṃ caiva sūtracchedabhayāt tataḥ //92// pūjāṃ kr̥tvā tathā homaṃ tatkarmachīdrakāraṇam / purāṇasūtram āpādya vibhajet tantujaṃ ca tat //93// badhvā pāṣāṇakhaṇde ca āgādhe‘mbasi nikṣipet / visvakṣeṇīyamantreṇa kṣiptvā ‘camya smared dharim //94// mūlamanttreṇa viprendra saptavārān samāhitaḥ.
1 A paper on the same subject was published as pp. 499–510 of the Pandit N.R. Bhatt Felicitation Volume, P-S. Filliozat, S.P. Narang and G.P. Bhatta, eds., Delhi, 1994.
2 On this dual aspect of mantras, see Hélène Brunner’s study ‘Mantras et mantras dans les Tantras śivaïtes’ in Le parole e i marmi. Studi in onore di Raniero Gnoli nel suo 70° compleano, Raffaele Torella et al., Rome: IIAO, 2001, pp. 183–212.
3 See Introduction, p. 95, and the next chapter.
4 On this tantra of the śākta Yāmala tradition (existing only in manuscript form), and on its commentaries, see A. Sanderson ‘The Sulva Exegesis of Kashmir’ in D. Goodall and A. Padoux, Mélanges tantriques à la mémoire d’Hélène Brunner, Pondichéry, 2007, pp. 231–424. On this tantra, and on the (perhaps thirteenth century) Mantramuktāvali, as well as on the ŚT, see also T. Goudriaan and S. Gupta, Hindu Tantric and Śākta Literature, Wiesbaden, 1981, pp. 49, 134–137,
5 1.75–78 refers to the Tantrik Texts edition of the TRT (reprint Delhi, 1981); 1.73–81 to the edition of its first chapter with the Sudarśana commentary of Prāṇamañjarī (Calcutta, 1940). This Tantrarāja is a different text from the tantra quoted by Abhinavagupta in the Tantrāloka as Tantrarājabhaṭṭāraka, which is the 24.000 śloka Jayadrathayāmala.
6 The absence of aṃśa is a serious defect since, as we have seen previously (see chapter 2, pp. 19–21, on mantravicāra, a mantra can only be made use of by a sādhaka if there is an aṃśa.
7 This is a different case from that of a mantra which is ‘asleep’ without being defective, and is to be ‘awakened’ before being made use of.
8 See Introduction, pp. 3–4.
9 On these practices, see the next chapter.
10 According to the Śivārcanacandrikā, the mantra in such a case is purified by repeating it thrice whilst raising the kuṇḋalinī and then doing a pūjā to the devatā.
11 See Agnipurāṇa, chapter 292 or Tārābhaktisudhārṇava (p. 324) quoting the Rudrayāmala.
12 This tantra is perhaps the same as the Mahocchuṣma quoted by Jayaratha in his Viveka on TĀ 1.18 as one of the eight Brahmayāmalatantras (ibid. vol. 1, p. 42), a group of texts where Kapālīśabhairava has Mahocchuṣmā among the four goddesses of his retinue (see A. Sanderson, ‘Saivism and the Tantric traditions’, in C. Carrithers et al. eds, The Category of the Person, London, 1988, pp. 660–704).
13 It is unclear how this is to be done.
14 On this term, see chapter 7, p. 96.
15 The term mātr̥kā for the letters of the alphabet is used when their nature of powers (śakti) is to be underlined – for instance, when they are ritually placed on the body by nyāsa.
1 The author is grateful to the authorities of the Kuppuswami Research Institute in Chennai for permission to reprint this article which first appeared in vols. LVI–LXII, 1986–92, of the Journal of Oriental Research, Madras, 1992 (pp. 65–76). This version includes, when necessary, a few corrections and emendations to the original text; the references given in the notes have, when necessary, been updated.
2 The NT, with Kṣemarāja’s Uddyota, was published as vols. 16 and 61 of the Kashmir Series of Texts and Studies (KSTS), Bombay, 1926–1939. This edition is reprinted in one volume edited with an introduction by V.V. Dvivedi, Delhi: Parimal Publications, 1985.
3 On this practice, see the entry on pratyaṅgira of the Tāntrikābhidhānakośa (TAK), vol. 3 (forthcoming).
4 These nine practices are as follows: dīpana, bodhana, tāḋana, abhiṣecana, vimalikaraṇa, indhananiveśana, santarpaṇa, guptībhāva and āpyāya. This list is partly the same as that of the usual ten mantrasaṃskāras used to counteract the mantradoṣas quoted, for instance, in the Śāradātilaka 2.112–123, or in the Kuārṇavatantra: on the subject, see here chapter 6.
5 The action of encasing is sampuṭīkaraṇa/kr̥iti, a term used more often, I believe, metaphorically, especially in the Trika, to describe the mutual encasement (that is, the conjunction or fusion) of two metaphysical entities, usually Śiva and Śakti. It is also used to express the mutual inherence of Śiva and the cosmos (see MVT 11.07; TĀ 3.205–206; Parātriṃśikāvivaraṇa, pp. 264–265, etc.). The term is used in the NṢA (1.29–30) to describe the coincidence of the first two (inner) triangles of the śrīcakra: one identified with Śiva, the other with Śakti.
6 These two works are the same text under two different names: Vāmakeśvarīmataa, with the commentary of Rājanaka Jayaratha, KSTS, LXVI, Srinagar, 1945. Nityāṣoḋaśikārṇava, with two commentaries, R̥ijuvimaśinī of Śivānanda and Artharatnāvalī of Vidyānanda, edited by V.V. Dvidedi, Varanasi, 1968.
7 This is a magical practice: cakra, bīja and letters are to be inscribed on gold leaf to be worn as a protective amulet (rakṣā).
8 The meaning of grasta in this context is not to be linked with its meaning in grammar/phonetics, where it is a fault of pronunciation due to the utterance of a letter hindered or held back at the throat – cf.R̥kprātiśakya XIV,3: jihvāmūlavigrahe grastam etat.
9 In their commentaries on NṢA or VMT, Śivānanda and Vidyānanda explain: aṅkuśaḥ kromkāraḥ.
10 This is clearly stated in Prāṇamañjarī’s commentary on TRT 1.72 where she describes several practices and says that there are still other ones. Cf. TRT with the commentary Sudarśana of Prāṇamañjarī critically edited for the first time by Jatindrabimal Chaudhury, Calcutta, 1940. Prāṇamañjarī lived in Varanasi in the eighteenth century.
11 See ‘Un terme technique de mantraśāstra, vidarbha’, JAs, CCLXV (1977), pp. 345–349.
12 These are usually pallava, yoga, rodha, grathana, sampuṭa and vidarbha (order of the PheṭK). The longer list of the NT originates perhaps – like the one previously given of nine mantric practices – in the Ucchuṣmatantra, that is, in an earlier stratum of Tantric texts.
13 In the PTV, p. 243 – see above, Introduction, p. 8 and note 17.
14 On the nature of mantras as well as on the non-dualistic Śaiva conceptions of the word, refer to my Vāc. The Concept of the Word in Selected Hindu Tantras (Albany: SUNY Press, 1990; Indian edition, Delhi: Satguru Publications, 1992), and to H.P. Alper, ed., Understanding Mantras (Albany: SUNY Press, 1988), which includes an – at that date – exhaustive bibliography on the subject.
15 On nyāsa, see here chapter 4.
16 The YH with Amr̥tānanda’s commentary, the Dīpikā, was edited by V.V. Dvivedi, Delhi: Motilal Banarsidass, 1998. I refer here to this edition which is the only reliable Indian edition of this work. French translation with introduction and notes by André Padoux: Le coeur de la Yoginī, Yoginīhr̥daya, avec le commentaire Dīpikā d’Amr̥tānanda, Paris: Collège de France, 1994.
17 An explanatory note with a drawing of the kāmakalā is given on p. 202 of my translation of the YH. On the kāmakalā, and especially its sexual connotations, see David G. White, Kiss of the Yoginī. ‘Tantric Sex’ in its South Asian Context, Chicago: The University of Chicago Press, 2000.
18 We saw above (p. 97) a nyāsa, on the image of the goddess Tripurasundarī, of her mūlavidyā, aṅkuśena vidarbhitā: intersected and bound up with the aṅkuśabīja KROṂ.
19 This is Sarasvatī, also called Vāgīṣī. The name Lipidevī, however, implies that this is a written alphabet goddess. Sarasvatī, in fact, holds a book in one of her hands; the written word is thus associated with her, on the divine plane.
20 It is described in several tantras, the Siddhayogeśvarīṃata, Mālinīvijayottara, Kubjikāmata, in the PTV, as well as in the AgP 145.
21 On the Mālinī, see the ṇādiphānta entry of the TAK, vol. 3. An interpretive study of this letter-deity is that of Somdev Vasudeva, ‘Synaesthetic Iconography: 1. The N. ādiphāntakrama’, which quotes several texts, published in D. Goodall and A. Padoux, eds., Mélanges tantriques à la mémoire d’Hélène Brunner/Tantric Studies in Memory of Hélène Brunner, Pondichéry: Institut Français de Pondichéry, 2007, pp. 517–550.
22 This is described in Jayaratha’s commentary on TĀ 3.67. On this, see my Vāc (quoted above, note 14), pp. 241–243.
23 On these diagrams, see the appendix of J.A. Schoterman’s edition of the Ṣaṭsāhasrasaṃhitā, op cit.
24 On mantroddhāra and the quest for aṃśa, I wrote an article published in the BEFEO 65, 1978, an English (revised) version of which is published here as chapter 2.
25 On should perhaps underline in this respect that the same Sanskrit term, pāṭha, is used for reading and for reciting. The coalescence of these two meanings is interesting, and is surely not accidental.
26 Transposed in another civilisation where the written sign is fundamental, mantras have evolved differently, as instanced by the siddham script of the Japanese Shingon sect, where Sanskrit bījas are used in meditation very much like Chinese characters. See on this R.H. van Gulik, Siddham. An Essay on the History of Sanskrit Studies in China and Japan, New Delhi: Sharada Rani, 1980.
ADDENDA 2009: On the nature of mantras and especially on the oral/written relationship, see A. Padoux, ‘L’oral et l’écrit. mantra et mantraśāstra’, translated here as chapter 9.
On the role of writing in India, see Charles Malamoud, ‘Noirceur de l’écriture. Remarques sur un thème littéraire de l’Inde ancienne’, in Paroles à dire, paroles à écrire, Paris: Editions de l’EHESS, 1997, pp. 85–114.
1 In this paper I take up again, but more briefly, a subject on which I wrote an article
‘Corps et mantra. De la présence des mantras dans le corps’ published in Oscar Botto, et al, eds., Du corps humain, au carrefour de plusieurs savoirs en Inde. Mélanges offerts à Arion Roṣu par ses collègues et ses amis à l’occasion de son 80e anniversaire, Bucarest et Paris, 2004, pp.563–578.
2 On this important notion, see Paul Schilder’s classical work on the subject: The Image and Appearance of the Human Body, first published in 1935.
3 One says, too, that the yogin attains a state of mantra: mantratvam.
4 This important text of the Pāñcarātra contains many ritual prescriptions. It was edited in the Gaekwad Oriental Series (vol. 54), Baroda, 1967. See on this Marion Rastelli, Philosophisch-theologische Grundanschauungen des Jayākhyasaṃhitā. mit einer Darstellung des Täglichen Rituals, Vienna, Verlag der Österreichischen Akademie der Wissenschaften, 1999. Though very interesting, this text has not been translated.
5 This mental worship is studied and carefully commented upon in chapter 12 of the thesis of M. Rastelli, quoted in the preceding note.
6 There are, as far as I know, no serious systematic studies of Hindu mudrās used in ritual. On a particular Tantric case, see A. Padoux, ‘The body in Tantric ritual: the case of the mudrās’, in T. Goudriaan, ed., The Sanskrit Tradition and Tantrism, Leiden, 1990, pp. 60–75.
7 This assertion is to be qualified, for a deity can very well be conceived as being ‘made of’ letters only. The mantra-god Navātman, for instance, is made up of the nine phonemes HSKŚMLVRYŪṂ, A Tantric deity having always one or several ‘faces’ (vaktra) and ‘members’ (aṅga), its mantric form will also include theses additions, which will also be made up of phonemes – see TĀ 30.15–16a. Are we faced, then, with mental images or with mentally evoked phonetic enunciations? As for Navātman, this god is also described as a beautiful young man, associated with different goddesses, notably the hunchbacked goddess Kubjikā: what is evoked in his case? The anthropomorphic figure, or the syllabic enunciation? On the navātman mantra, see the navātman entry of the TĀK, vol. 3.
8 On this, see the Somaśambupaddhati, vol. 3, edited and translated in French by Hélène Brunner (Pondicherry, 1977), on the Śaiva initiations. Plates in the appendix to that book show how such cosmic elements are placed on the body of the initiand.
9 Such, for instance, as those illustrating the Ṣaṭcakranirūpaṇa edited by A. Avalon as The Serpent Power, being the Shat-Chakra-Nirūpana and Pādukāpañcaka, Madras: Ganesh & Co., 1953.
10 The process is described in full detail by M. Rastelli, in pp. 246–270 of her study Philosophisch-theologische Grundanschauungen der Jayākhyāsaṃhitā (Vienna, 1999).
11 This mental rite is studied by A. Sanderson: ‘Meaning in Tantric ritual’, in A.M. Blondeau and Fr. Schipper, eds., Essais sur le rituel III, Paris: Bibliothèque des Hautes Etudes, Sciences Religieuses, vol. CII, pp. 15–95.
12 This worship of the goddesses is described and commented by A. Sanderson, ‘Maṇḋala and Āgamic Identity in the Trika of Kashmir’, in A. Padoux, ed., Mantras et diagrammes rituels dans l’hindouisme (Paris 1986), with a drawing showing the body of the adept drawn by Stefanie Sanderson – printed again on p. 308 of the 1999 reprint of R. Gnoli’s translation of the TĀ (Milano: Adelphi, 1999).
13 I am not sure ‘lived body’ is good English – probably not. What I wish to refer to here is the distinction put forward by Max Scheler (and taken over by existential psychoanalysis, the Daseinanalyse) between the Leib, the body as it is experienced, lived, and the Körper, which is the perceived, physical, anatomic body; both of course are never separated – one does not ‘live’ merely with one’s physical body.
14 On uccāra, see the entry on this word in vol.1 of the TĀK. prāṇa is not the respiratory breath, but a ‘subtle’ inner vital ‘breath’. L. Renou defined it as ‘l’aspect vital de l’ātman’. On this subject, see the entry prāṇa of the TĀK.
15 In S. Gupta, D.J Hoens, T. Goudriaan, Hindu Tantrism (Leiden, 1979). Though purely imaginary, the yogic body, being ‘intraposed’ within the anatomic body, is both very largely delineated on it and linked to it when its aspect and functioning are visualised: all this ‘happens in the body’. This authorises some – Mircea Eliade, for instance – to speak in such a case of ‘physiologie mystique’: the expression is debatable, but it underlines well the link, the interaction of those two ‘bodies’.
16 This important tantra is often quoted or referred to by Abhinavagupta. It deals mostly with the cult of mantra-goddesses possessing magical powers. A shorter version of this work was edited and translated in English by Judit Törzsök for an Oxford (unpublished) thesis. A fuller edition is now expected.
17 The SvT describes this in the cosmic and soteriological perspective of the six ‘ways’ (adhvan).
18 The NT completes the series of twelve kalās with those from nādānta to unmanā as corresponding to different, ever higher, planes of the deity. A precise and very usefully commented description of this passage of the NT is given by Hélène Brunner in her study of that tantra: ‘Un tantra du nord: le Netratantra’, BEFEO, LXI (1974), pp. 125–197.
kūṭatraye mahādevi kuṇḋalinītraye ‘pi ca/
cakranāṃ pūrvapūrveṣāṃ nādarūpeṇa yojanam//
20 For the Yoginīhr̥daīya, as for many Śaiva texts, the kuṇḋalinī is considered as comprising three sections: igneous, solar and lunar. The first, igneous, section starting from the lowest cakra, the second part, solar, extends from the heart-cakra to the bhrūmadhya, where begins the third one. To this threefold division corresponds that of the śrīvidyā; it is also present in the nine cakras constituting the śrīcakra which are also distributed into three groups. During the japa, all these three groups are brought together by the ascent of the nāda which prolongs the uccāra of the three HRĪṂ.
21 For the Śaiva cultic practice as described in the āgamas (and as it is still, in some cases, performed in India today), see R.H. Davis, Ritual in an Oscillating Universe. Worshipping Śiva in Medieval India, Princeton: Princeton University Press, 1991. We, of course, do not know if a real mystical union with the deity in ritual was current among Śaiva adepts in the tenth to fourteenth centuries. We know, on the other hand, that officiants often limit themselves to the mere enunciation of the prescribed mantras – mantraprayoga – without trying to do anything else than follow formally the prescriptions of the ritual handbooks: this is indeed very far from the world of the ancient texts – but we do not know much about that world.
22 The big toe of the right foot is the spot where Kālāgni, the (destructive) Fire of Time which burns the cosmos to cinders, comes forth. We see here that the destruction of the body of the adept is foreseen.
23 The cakras or granthis, spots, ‘nods’ where the power, the prāṇa, concentrates are in a way obstacles for they are thresholds to be passed. This is why, to progress, they are to be cut or pierced: granthiccheda. See the entries on these terms in the TĀK, vol. 2.
24 On khecara and the state of being khecara (khecaratā/khecaratva), see the corresponding entries of the TĀK, vol. 2,
1 Translation of ‘L’oral et l’ecrit. Mantra et mantraśāstra’, in C. Champion, ed., Traditions orales dans le monde indien, Paris: Editions de l’EHESS, 1996, pp. 133–148.
2 They are, in fact, deities: Tantric pantheons are pantheons of mantras.
3 On the ‘circulation’ of a mantra in the nāḋīs, see, for instance, chapter 7 of the NT which describes the ‘subtle meditation’ (sūkṣmadhyāna) of the Netramantra OṂ JUṂ SAḤ. This mantric practice is briefly described in Brunner, 1974, and Padoux, 1990, pp. 411–406.
4 The recordings on cassettes (or otherwise) of the mantras of the pūjā that one finds nowadays in India show another aspect of the ‘vulgarisation’ of mantras. The ritual formulas of the usual daily Hindu worship, however, are not secret – differing in this from those for the worship of esoteric Tantric deities.
5 We must not forget, on this subject, that silent, purely visual, reading is a comparatively modern usage. Earlier, reading was pronouncing, thus giving back to the oral its rightful place. See Ong, 1982.
6 uddhāra: from ud-DHR̥, take out, raise up, select, extract, See, here, chapter 2 on the extraction and selection of mantras.
7 On prastaras and gahvaras, see J.G. Schoterman, The Satsahasrasamhita, Brill, 1982.
8 On the rites through which an adept – a sādhaka – masters the mantra he has received, see Brunner 1974a, pp. 411–443.
9 On these rites, see here chapter 2, pp. 18–22.
10 That is, the imaginary structure of centres (cakra) and channels (nāḋī) conceived as present in the physical body. It is usually, but wrongly, called ‘subtle body’; the subtle body (sūkṣma- or liṅgaśarīra) properly so-called being something else.
11 On this sense of dhyāna, see the entry on that word in the TAK, vol. 3 which refers to G. Bühneman, The Iconography of Hindu Tantric Deities (two vol., Groningen. Egbert Forsten, 2000–2001). By nyāsa one may also place divine entities, cosmic divisions, yantras or cakras, etc.
12 When tantras, etc., name a mantra, they generally quote the r̥ṣi who has first ‘seen’ it, its metre (chandas) – which gives it a ‘vedic’ air – then the deity (devatā) it expresses, and the syllable which is its seed (bīja). To these are sometimes added the power (śakti), the point (kīlaka) and the use (viniyoga) of the mantra.
13 A triangle (point down) is the symbol of the goddess, of the feminine. The phoneme E is thus in essence the locus of creation – and of bliss. It is, Jayaratha says (ad, TĀ 3.94), the ‘mouth of the Yoginī’ (yoginīvaktra) where one receives initiation and experiences the bliss of liberation. But vaktra, in this case, is not the mouth, but the sexual organ of the Yoginī, where the yogin in sexual union with his ritual partner attains bliss and fusion with the Absolute. The deity, in this case, is above all form, whilst being also (at least metaphysically) triangular.
This sort of practice is described in chapter 29 of the Tantrāloka – on which see Silburn, Kundalinī. The Energy of the Depths, Albany, 1988.
14 On the uccāra of OṂ, see Padoux, 1990, chap. 7, pp. 402–411. On the kāmakalā, see Punyānanda’s Kāmakalāvilāsa, edited and translated by A. Avalon, or, for a study and interpretation of this diagram, David G. White, ‘Transformation in the Art of Love. Kāmakalā Practices in Hindu Tantric and Kaula Traditions,’ History of Religions, 18/2, 1998, pp. 172–198.
15 Buddhist dhāraṇis were translated in Chinese, written therefore since the fourth century. But the cult and homage paid to the sūtras as holy ‘Scripture’ seem to have existed earlier in the Mahāyāna traditions. The most marked cult of the Book is that of the Sikh Adigranth, but it is by reciting it that it is honoured and that a beneficial result is produced. See Dusenbery 1992.
In the Hindu field, a purāṇa, as a book or, better, in manuscript, may be honoured. Such is the case notably of the Devīmāhatmya: see Coburn, 1991.
16 On the mantiravatis, see Diehl, 1956.
17 On the anārthakya (lack of meaning) of mantras, see Renou, 1960, specially, pp. 70ff.