A hall mark feature of the Hindu tradition right from its proto-Indo-Iranian antecedents has been the mantra shAstra. It is the mainstay of both the orthopraxy (AstIka matas) as well as most heteropraxies (namely bauddha and jaina matas) of the dharma. The technical basis of mantrashAstra (a secret only understood by practicing insiders with a knowledge of the texts or shAstra-s learned with assistance from an AchArya and experienced with assistance from a guru) has been remarkably conservative and stable since its earliest traceable history going back to the proto-Indo-Iranian period. It is on account of this stability of its foundational basis that many traditional adepts are not very particular in distinguishing the various historical layers of the mantra-shAstra. Thus, they exhibit a fluid syncretic tendency of melding different layers and traditions. Unfortunately, even in the modern era few adepts have acquired education in evolutionary analysis and Indo-European and Dravidian linguistics to produce hypothesis to analyze the historical developments in mantrashAstra effectively. Yet, even a cursory glance of the mantrashAstra illustrates two great categories within it: 1) the older vedic mantrashAstra and 2) the later tantric mantra shAstra. Nevertheless, the latter exhibits many features which appear to have been derived from seeds in the former.
The vedic mantras probably right from the PI-Ir exhibit a diversification into 3 distinct forms simultaneously deployed in major shrauta rituals — namely the R^iks, yajushes and sAman-s. The R^iks where composed in Chandas and declare their R^iShi, Chandas and devatA with great care. Special deployments like the sAmidhenI-s or kathayanta prayoga or the nyUnka and ninArda style recitations were often decorated with “bIja-s” like OM, hiM ShoM, ShomsAvOM and akShara-s like su, mat, pad, vag, de. These elements respectively are the precursors of the R^iShyAdi nyAsa of the tantric mantras and resemble the insertion of bIja-s in mantra-s of the tantric mantra shAstra. The yajushes are often ritual formulae with elements like namaH, vauShaT, vaShaT and svAhA (also phaT as per yajurvedic, atharvanic and sAmavedic tradition), which also comprise the typical elements of most major tantric mantras. The taittirIya AraNyaka while laying out the abhichAra formulae for the deadly forms of agni and the” killing by the year” formula gives the following incantations to be used with them(TA 4.27.1) :
khaT phaT jahi | ChindhI bindhI handhI kaT | iti vAchaH krUrANi
i.e. the calls KhaT phaT and jahi (slay) , ChindhI (rend apart), bindhi (split up) , handhi (smite) and kaT are used in fierce (i.e. abhichAra) rites.
Again the taittirIya AraNyaka supplies similar calls in the formula called “the placing in yama’s jaws” (TA 4.37.1):
yadamuM yamasya jambhayoH | AdadhAmi tatha hi tat | khaN-phaN-mrasi ||
The phaT, jahi chindhi etc calls seen in these yajushes are typical of later tantric mantras with similar intent.
The magical stobha-s in the mantra-s of the sAmaveda also bear a resemblance to the bIja-s of the later tantric mantra shAstra.
However, the divide between the two is still apparent:
1) The tantric mantras are largely dominated by bIja-s and standalone non-semantic akShara clusters, whereas the vedic mantra-s are predominantly semantic in structure with the exception of the musical sAman where the underlying R^ik is greatly distorted (except certain stobha only songs, which lack a R^ik). Most vedic mantra-s embed a lot in the meaning of “sentences” and as a result are typically longer than a tantric mantra, whose semantically meaningful elements are very short (e.g. sudarshana chakra mantra: OM sahasrAra huM phaT; sahasrAra= thousand spokes – that is all ! ) or may even be non-existent (e.g. the shrIvidyA kAdimata: ka e i la hrIM | ha sa ka ha la hrIM | sa ka la hrIM||).
2) The vedic canon is very well preserved with great fidelity and little admission of interpolation and syncretism of extraneous elements. In contrast, while the tantric mantra-shAstra admits to many “canonical” texts, these hardly exhibit anything close to the level of fidelity in vedic preservation, instead evolving rapidly while only preserving the mantra and intent. An extreme case being the rudra-yAmala that is taken as a canon but is currently a jumble of several texts including some questionable attributions.
3)The shrauta ritual is very standardized and provide a strict system of orthopraxis. The tantric rituals while admitting the concept of orthopraxis is more ready in admitting syncretic elements as a result of which different traditions exhibit much greater melding.
Now the big evolutionary question is: how did the transition occur from the vedic to the tantric form? The gR^ihya rites with one fire and a sthaNDila and the early vidhAna rituals like those in the R^igvidhAna might offer clues in terms of the praxis. But that is not the issue I am seeking to explore here. I am primarily concerning myself about the mantras themselves.
In this regard one might note the following disparate elements that might offer clues:
1) The early post-vedic compositions like nAma-saMgraha-s for devatA-s (e.g. the celebrated viShNu sahasranAmam from the mahAbhArata). These appear to be predecessors of the sahasra and shata nAmAvalI-s that proliferate in the tantric period and form a major component of tantric worship, even if not as the primary mantras.
2) Not all tantric mantra-s are bIja mantra-s but a small subset of them are actually anuShTubh-s. One of the potentially oldest and most famous of these is the ugra viShNu or mantrarAja anuShTubh (ugraM vIraM mahAviShNuM jvalantaM sarvatomukhaM | nR^isimhaM bhIShaNaM bhadraM mR^ityu-mR^ityuM namAmyahaM ||)
A collection of such anuShTubh mantras include: 1) ugra viShNu 2) the kArtavIryArjuna naShTa dravya lAbha mantra 3) vana-durga mantra 4) the shAstA mantra which identifies him with revanta. 5) The skanda anuShTubh mantra. Also related to this point are the gAyatrI mantras for specific deities that form an important, though not central aspect of tantric worship. These gAyatrI-s with the 3 pada-s bearing the words “vidmahe”, “dhimahi” and “prachodayat” with 3 names of the deity are all modeled after the rudra gayatrI-s found in some kR^iShNa yajurvedic saMhitA-s, which in turn is modeled after the R^igvedic savitrI of vishvAmitra.
3) Mantras associated with circum-vedic traditions, but distinct from traditional vedic and tantric mantra-s: E.g. the AsurI kalpa, the uchChuShma kalpa and dhUrta kalpa of the atharva veda parishiShTha which provide mantras and rites for the worship of AsurI durgA, uchChuShma rudra and kumAra. It is this class of mantras that might offer the important bridge between the tantric mantras and yajushes which as noted above have elements similar to the tantric formulae.
Of these elements the concept of nAmAvalI-s appear to go back to the proto-IIr period– in the Iranian tradition we encounter the 100 names of ahurA-mAzdhA that clearly resemble the later nAmAvalIs of Indic tradition. While there are no explicit nAmAvalI-s in the veda, the vedic namaH mantras of the yajur (e.g. the shatarudrIya) appear to be an archaic relative of the nAmAvalI. This become apparent when one compares it with the early-post-vedic hymn of dakSha to rudra. Thus, the nAmAvalI mantras appears to be an early concept in Indo-Iranian worship that appears to have been merely expanded greatly in the tantric realm.
The anuShTubh mantras clearly differ from all tantric mantras both in lacking bIjas and in their form, which resembles R^ik-s. Some of them like the mantra-rAja are clearly fairly ancient. We also observe such mantra-s in the pre-tantric circum-vedic texts like those of the pAshupata-s (e.g. rudraM kruddhAshanimukhaM devAnAM IshvaraM paraM | shveta-pi~NgalaM devesham prapadye sharaNAgataH ||). These features suggest that the evolution of the mantra shAstra passed through a phase where new R^ik like mantra-s, but mainly in the shloka meter were composed. This stage of shloka mantra-s might have also overlapped with the vidhAna period and the deployment of R^iks in a tantricized context. For example the well-known yAm kalpayanti R^ik of pratichIna a~Ngirasa and the indra mantra “yata indra bhayAmahe…” are deployed in later AV tradition in a tantric form.
But it is in the parishiShTa kalpa-s of the AV that one finds the true precursors of the tantric mantras. In a sense these mantra-s are descendants of vedic mantras such as those of the arrows of rudra found in the AV and YV traditions (e.g. AV-shaunaka 3.26 and 3.27).
Examples of abhichArAdi mantras to the uchChuShma rudras from the uchChuShma kalpa:
kAlAya karAlAya namaH svAhA | (shatru nAshanaM)
amoghAya namaH svAhA | (shatru AveshanaM)
vR^iShabhAya namaH svAhA | (shatrUnAm dhanakShyaM)
uchChuShmAya namaH svAhA | (unmAdanaM)
uchChuShma-rudrAya namaH svAhA | (shatru dahanaM)
aghorAya namaH svAhA | (roga nivAraNaM)
suvarNachuDAya namaH svAhA | (roga abhibAdhanaM)
hAhAhIhI namaH svAhA | (go lAbha mantraM)
mahAvaktrAya pi~NgalanetrAya namaH svAhA | (vR^iShTi)
khanakhanAya namaH svAhA | (prasiddhi prApti)
ghanaghanAya namaH svAhA | (yuddha vijaya)
mahApashupataye namaH svAhA | (pashubhyo .abhirakShaNaM; from animals)
Other examples of longer mantras from the uchChuShma kalpa are:
namaH kaTa vikaTa kaNTe mATe pATale vikale asauryAsau asauryAsau pR^ithivIShTakA iShTakAjinAtyUnyo saugaluM tigaluM te kaTam asi kaTapravR^ite pradviSha rudra raudreNAveshayAveshaya hana hana daha daha pacha pacha matha matha vidhvaMsaya vidhvaMsaya vishveshvara yogeshvara maheshvara namaste .astu mA mA hiMsIH huM phaT namaH svAhA || (a japA mantra)
shive jaTile brahmachAriNi stambhani jambhani mohani huM phaT namaH svAhA || (an Atma-rakShA mantra)
Interestingly, two of the mantras from one of the earliest tantric texts of the bauddha tradition known as the mahAmAyUrI-vidyA-rAj~nI are orthologous to the above two longer mantras of the AV kalpa.
The AsurI kalpa mantra (shaunaka version):
oM kaTuke kaTuka-pattre subhage Asuri rakte raktavAsase atharvaNasya duhitre aghore aghora-karma-kArike [amukaM] hana hana dahadaha pacha pacha matha matha tAvad daha tAvat pacha yAvan me vasham Anayasi svAhA ||
The AsurI kalpa mantra (paippalAda version):
oM namaH kaTuke kaTuka-pattre subhage Asuri raktavAsase atharvaNasya duhitre ghore ghora-karmANi-kArike [amukasya] prasthitasya matim bhaja upaviShTasya bhagaM daha shayitasya mano daha pravR^iddhasya hR^idayaM daha daha daha tAvad daha daha pacha pacha matha matha tAvat pacha yAvan me vasham AgachaChet svAhA ||
The shorter atharvanic kalpa mantra-s resemble the shorter mantra-s typical of the later tantra-s, while the longer ones from both AsurI and uchChuShma are closer to the pada-mAla mantras of later tantric mantra-s. The mantras of the AV kalpa share certain specific major features of the tantric mantras:
1) Use of imperative terms like hana, daha and pacha. These words are already seen in the AV saMhitA mantras. e.g.:
Chindhy A Chindhi pra Chindhy api kShApaya kShApaya || (AV-S 12.5.51)
vR^ishcha pra vR^ishcha saM vR^ishcha daha pra daha saM daha || (AV-S 12.5.62)
Imperatives in mantra are seen right from the earliest layers of the vedic mantras i.e. from the RV itself. They increase some what in prominence in the AV tradition (see darbha sUktaM and are extremely prevalent in tantric mantras, especially the mAlA mantras. An ancient example of such a mantra, albeit in the form of R^ik is RV 10.87.23:
“viSheNa bha~NgurAvataH prati Shma rakShaso daha |”
With poison turned against the vicious rakshas burn away!
2) onomatopoeic forms like hAhA hIhI and 3) the invocation of shaktis as seen in the AtmarakShA mantra of the uchChuShma kalpa and the AsurI mantra of the AsurI kalpa.
However, these kalpa mantras are very light on bIja-s, at best using the form OM and huM already known from vedic tradition. Thus, in the AV kalpa-s we see the ancestors of tantric mantras and connect them to their vedic mantra precursors. This also suggests that the major transition into classical tantric mantra-shAstra was accompanied by the innovation of the extensive use of bIja-s (an atavistic expression of the vedic linguistic tradition in a new form?).