It is nearly inconceivable for the average modern Hindu to imagine his religion as being aniconic. Worship at temples is an average Hindu’s major or only of religious activity, and it is not uncommon for practically every Hindu to keep some household images or pictures of deities. Yet, there is hardly any sign of temples or worship in temples in the vedic saMhitA-s, which are dominated by fire and soma rituals. Even in the itihAsa-s, while temples are mentioned several times, their description is not given much prominence. For example, in the rAmAyana (5.43.1-5) hanumAn is mentioned as breaking a large and lofty temple in lankA, but no other specific details of it are provided. The temple in the nikumbhilA grove in lankA is mentioned as a spot where meghanAda was to perform a fire rite for invincibility, but again hardly any details of this temple are mentioned. In the great bhArata too temples are mentioned, but are hardly prominent in its enormous bulk that otherwise covers an extraordinary diversity of topics. In itihAsas as a whole, references to temples, typically termed chaitya, are an incidental element in the context of describing holy tIrtha-s or urban centers (see below). The earliest available works in the dravidian languages, the tamiL puranAnuru and akanAnuru also do not display any notable dilation on temples, despite having several religious themes in them. The mAnava dharma shAstra and other early dharma works too are either silent or limited in their allusions to temples. In fact, the MDS recommends against brAhmaNa-s who earn their living as temple priests from being employed as a sacrificial invitees (3.152) or as recipients of brahminical donations (3.180).
Thus, it would appear that the foundational texts of the Hindu dharma, had few if any references to temples, and these played a relatively minor role in the early phase of the dharma. This is in striking contrast with the extant purANas and the modern Hindu culture where temples, temple construction, iconography and iconic installations play a prominent role. Thus, it almost appears that there was a major memetic shift in the Hindu worldview at some point in history which marked the transition to this more “modern” form of temple-centric worship.
Such a view is not inconsistent with archaeology: Even the highly urbanized Indus civilization and associated chalcolithic cultures of India (leaving aside the dispute of whether or not they had Aryan components), with a whole range of constructions and technologies, show little evidence of prominent temples comparable to those seen in medieval India or contemporary Egypt or Mesopotamia. Again, after the collapse of the urban Indus civilization, we find large classical temples emerging relatively late in course of the second urbanization of India. So what are the origins of temples in India? Related to this question is the parallel question: What were the origins of iconic worship in India?
Many branches of the Indo-European tree show signs for veneration of icons: Greeks, the syncretic Neoplatonic religion, Romans, Germanics, Balts and Slavs, Iranians and Indians. But not all Indo-European groups show evidence for large-scale temple construction right from inception: we find little evidence (as we saw above) amongst early Indians, early Balts or even Iranians. So it is not impossible that iconic worship of some kind was always there amongst Indo-Europeans from the ancestral period, but construction of public temples was potentially a convergent trait. We do hear zarathuShTra the founder of the Iranian cult of mAzdhayasna ranting and raving against icons:
“kadA azem mUrthem ahya maghahya yA Angraya Karpanao urupayeinti”
rendered in Sanskrit would be:
“kadA aham mUrtim asya maghasya yA A~Ngirasa karpaNA ropayanti”
When [ahurA-mAzdhA] can I uproot the idol from this assembly, that set up by the angra-s and the karpaNas?
That such icon-worship possibly persisted in conflict with the zarathuShTrian cult in the Iranian sphere is suggested in the boast of the Iranian emperor Xerxes (a follower of zarathuShTra’s cult): “I destroyed this temple of daevas”. As an aside it possible that the marriage and political relationships of these Iranians transmitted such iconoclastic memes to Afro-Asiatic cultures lying further west in the Middle East, which then formed a basic tenet of the Abrahamisms. In any case these observations would also support the contention that icon worship, even if not in temples, but in congregations or assemblies, like those mentioned by zarathuShTra as being set up by the a~Ngirasa-s, did exist among the early Indo-Iranians.
With this in mind we shall next seek traces of such icon worship and connections to the origin of Hindu temples if any.
chaitya: The earliest word used for temple in the Hindu world is chaitya. This word is encountered in the gR^ihya sUtra-s, AV parishiShTha-s, and itihAsa-s. Its etymology is simple: piled up, from the root chit to piling. Now, in the vedic language chit is often used in a specific sense, i.e. piling up of the fire altar or agni-chiti (as in agni-chayana) from bricks. This suggests that the word chaitya similar implied a piling up bricks to form a shrine. This is consistent with the observation that earliest temples of India are relatively simple piled brick structures. We also encounter the word chaitya vR^iksha to describe 4 species of large fig trees in the late vedic texts and itihAsa-s. From the earliest depictions to modern times we find these figs (most often ashvattha or audumbara) being shown with a platform of bricks piled around their base. This hints the form of the earliest public shrines in India: piled brick structures, perhaps even at the base of large fig species. Consistent with this, the sparseness of the description of chaitya-s in the itihAsa-s, makes it likely that the early chaitya-s were rather low-key structures rather than the grand temples of the purANa-s. In the MBh 3.125 (critical) it is mentioned by lomAsha to yudhiShThira that the tirtha on the Archika hill is a place where there are numerous (100s) chaitya-s for the 33 gods. Again in MBh 3.121 lomAsha mentions the chaityas on the banks of the narmada to be visited by the pANDu-s. Both the bhArata and the rAmAyana mention the presence of chaitya-s in graveyard using a very similar cliché as a part of a simile to describe something terrible.
MBh: shmashAna-chaitya-drumavad bhUShito.api bhaya~NkaraH |
Ram: shmashAna-chaitya pratimo bhUShito.api bhaya~NkaraH |
Even today graveyards often have temples of rudra. Another point of note is that in the MBh the term chaitya is often combined with yUpa: and the earth or a place is described as decorated by chaitya-s and yUpa-s (“chaityayUpA~NkitA bhUmi”). yUpa-s from shrauta rites were often left behind at the end of a yAga as a mark of their performance. Their linkage suggests that chaitya-s were possibly small shrines that might have been likewise associated with holy spots were rites were performed.
chaitya yAga: The gR^ihya-sUtra-s appear to be the earliest texts mentioning the term chaitya. An interesting instance is the mention of the rite termed the chaitya yAga in the AshvalAyana (1.12.1) and pAraskAra (3.11.10) gR^ihya sUtras. From what we can glean from these sUtra-s, the chaitya yAga involved making of fire oblations to the deity of the chaitya in the house of the worshiper. Then, before the final expiatory sviShTakR^it oblations, the worshiper offers food as bali to the deity of the chaitya. If the chaitya is far away the votary is advised to wrap two aliquots of the bali offering in a leaf and send it via an emissary to the chaitya. It specifically instructs that the votary gives the messenger food/payment, and if necessary a boat (that is if a river-crossing is involved) and weapons (if dangers are expected). At the face of it, the nikumbhilA rite of meghanAda in the rAmAyaNa might have been some form of a chaitya yAga. As we have discussed earlier the vaikhAnasa gR^ihya sUtra also mentions a rite at a shrine of kumAra in connection to a rite of travel for a young child. Hence, at least by the latter layers of the gR^ihya sUtra-s, there were public shrines that might be distant from where the votary is located. An enshrined deity may be worshiped using the standard vedic technique and the offerings (bali) from the rite may be carried to that deity in a specific shrine.
Early mentions of chaitya-s in nAstIka texts: To gain a perspective from outside of the AstIka stream within the dharma, we could turn to the two major nAstIka streams, whose origin was coeval with the latest layer of the vedic period. Their earliest texts, as well as the early biographies of their founders, the nirgrantha and shuddhodana-putra, mention several chaitya-s as being already present and worshiped by people. Both bauddha and jaina sources typically call them yakSha-chaitya, because of their wish to downgrade the ancestral Hindu deities relative to the founders of their cults. Yet it is clear that many of these were shrines of typical Hindus deities, whereas others of real yakSha-s. For example both bauddha and jaina sources mention a shrine of devI ShaShThI, the wife of kumAra being present at vishAla at the time of two nAstIka founders. The jaina sources also recall a chaitya of skanda at srAvastI in the days of mahAvIra and chaityas to gods like shUlapANI (rudra) and a large one of the yakSha pUrNabhadra (worshiped even today by jainas). The latter chaitya and that of ShaShThI were even visited by mahAvIra in his peregrinations. The graveyard chaitya might have also been behind the use of the term chetiya (a prAkR^ita form of chaitya) for the funerary relic shrines of the tathAgata and his successors worshiped by the bauddhas.
Thus, by the period of 600-450 BC already several chaitya-s to deities were in place throughout the urbanizing northern Gangetic belt. The available descriptions of such chaitya-s suggest that they were small to medium-sized constructions (usually brick) usually surrounded by groves of or under trees (typically figs). By the mauryan period described in the arthashAstra forts and cities had chaitya-s for a number of deva-s and devI-s like indra, ashvin-s, kumara, rudra, kubera, madirA and aparAjitA. A description of the chaitya of kaumArI suggests that it had multiple AvaraNa one inside the other and a circular arch. So we can see that between the time of the founders of the nAstIka-mata-s to the mauryan era the chaitya-s appear to have steadily increased in importance, becoming integral aspects of city life. However, there is still no hint that they were large structures like the classical Hindu temples.
Iconic worship of the late gR^ihya period and the rise of temple worship: Clues regarding the early forms of image worship come from specific references in what may be termed the late gR^ihya literature. These can be assigned to the later layer of gR^ihya sUtra literature because they do not appear to be conserved across all gR^ihya sUtra-s and are often found in what are called parishiShTa or sheSha sUtra-s. The most prominent examples of these are:
bodhAyana sheSha sUtra 2.13-15, viShNu-pratiShTha kalpa: gives the rite for the installation of an image of viShNu, its daily worship and its bathing.
vaikhAnasa smArta sUtra 4.10-12, viShNvarchana vidhi: gives a similar rite of installation and routine worship of an idol of viShNu.
bodhAyana sheSha sUtra 2.16-18, rudra-pratiShTha: kalpa gives the rite for the installation of an image of rudra, its daily worship and ablutions.
bodhAyana sheSha sUtra 4.2, dhUrta bali: mentions installation of an idol of kumAra in a maNTapa decorated with leaves.
atharva-veda parishiShTa, skanda yAga: mentions a very similar installation of a kumAra icon in a platform with leaf and mirror decorations.
The shAnti kalpa of the atharva-veda: mentions making images of the nava-graha-s their installation and detailed procedures of worship.
Key features of the majority of these installation rituals include: 1) The rites do not use prANa-pratiShTha mantras typical of the Agamic procedures 2) They might use an “eye-opening” rite using a golden needle with vedic mantra-s. 3) They do not use any classical tantric mantra-s of the Agama-s for the worship of the deity, instead using vedic and some new specific mantra-s specified by these gR^ihya texts.4) They always involve a vedic-styled homa with oblations made to the deity being installed.
This process epitomized by the vaikhAnasa viShNvarchana vidhi for the installation and worship of a small pratima of viShNu (6 a~Ngula-s in height), has been expanded by the later vaikhAnasa Agama-s for the installation of the viShNu idols in huge temples. Thus, we see a link between the gR^ihya image worship and the origins of the larger image worship in temples. However, in most streams except the vaikhAnasa-s, this shift to the large temples was closely linked to the rise of a new class of texts and rituals in place of the vedic ones – the Agamic texts. Thus, one might state with confidence the origin of the classical Agama-s as well as extant paurANic narratives are closely linked to the emergence of the large temples. It is not without reason that town-planning, civil constructions and the arts occupy the interest of early Agamas like devyAmata and purANa-s like viShNudharmottara.