A Novice Views India, Part II: Pallava Temples at Mahabalipuram

The second of my Indian photo posts has become a bit of a monster. In the first version I already had upwards of fifty pictures I wanted to share, which is probably excessive though it’s a fraction of what I took, but I tried to save space on commentary and just let the pictures speak for themselves. Then I talked about the chronology of the site concerned with someone, realised I had to do some actual reading and the result has wound up being a total rewrite that is huge in both graphic and textual content! But I think it’s worth it, because the site concerned is one of those places, like Montserrat, that I wish somehow that everyone could visit at least once. It is another plurifocal religious site in liminal space, too, but there the similarities end because where we’re talking about is the early medieval or late antique temple complex at Mahabalipuram in Tamil Nadu.¹

As you can see there, I’m already fudging questions of chronology, so let’s get that as sorted as it’s going to get by breaking this fairly complex site down a bit. Mahabalipuram, also known in its time as Māmallapuram or the Seven Pagodas, is a site with two visible foci, one down at the shoreline of the eastern Indian Ocean…

The Shore Temple complex, seen from a distance; we will come back to it closer up

… and one inland from there, with too much in it to get into one photo.

Therefore, here’s the shore zone again, but viewed from the site of the Olakkanesvara temple at the highest point of the inshore site, and we will also go round the inshore site in more detail in a minute

Inland from that, there’s the modern town which itself has a few old religious buildings in it, but that’s not what I’m talking about today, I just can’t.

This was actually our point of arrival and departure, and as you can see has some worthwhile old architecture of its own, but on the way in we were eager to get to the main site and by the way out, kind of overwhelmed, so it never got its due, sorry

Now, I said visible, because complicating all this is the fact that the same tsunami of 2004 that wrecked the Portuguese mission church looked at in the previous one of these posts also exposed much older structures here. Some sculptures are still visible and there are some underwater structures which are not, but which were spotted by presumably-terrified locals as the water drew back before the wave, and have since been dived upon in calmer times, but about which as yet very little is known. Those structures themselves probably went underwater in a different tsunami of 954.² So there’s probably a third focus, which might be oldest of all, but we don’t know. That last clause will be a running theme. But right now, two main foci, shoreline and inshore. OK?

The seafront, seen from the Shore Temple, and including one of the older sculptured rocks now exposed on the beach-front; but how much more is unexposed beneath the waves, eh?

OK. Now, there are also three broad categories of building at Mahabalipuram, as well as quite a lot of free-standing sculpture, and some things that are really neither of the above.

This rock is called Krishna’s Butterball, and contrary to appearances shows no signs of going anywhere any time soon, but whether it was carved out, put there or just fell or rolled, no-one knows

In the building categories, there are structures that were built from the ground up with materials transported to location (even if probably not very far);…

The Olakkanesvara gopura, seen from (well) below; the Shore Temples, to which again we will shortly come back, would also fit in this category

… there are structures which were made by carving out caves in the huge rock formations that stand above the current surface, though I do feel like ‘cave temples’ sells these pieces of architecture short;…

Frontage of the Rāmānuja Maṇḍapa, to which I’m not going to dedicate separate attention, sorry, but something had not to fit

… and there are structures that were just hewn bodily out of those rocks until there was only monolithic structure left and no rock around it.

South-west corner of the Arjuna Ratha, one of a set of five of these to which we will come back

Some of these buildings have inscriptions, but not all, and not all the inscriptions are firstly either particularly helpful…

This is apparently the only inscription I noticed, and it’s not a great photograph, sorry. Of course, I can’t read it either, which doesn’t help, but if I’m right it’s one of four instances on the site of the so-called Śaivite Curse, translating roughly as, "Six times cursed be those in whose hearts does not dwell Rudra, the deliverer from walking on the evil path!". It’s on the floor of the Rāmānuja Maṇḍapa, but obviously doesn’t help a great deal with identifying its patron! (See Marilyn Hirsh, "Mahendravarman I Pallava: Artist and Patron of Māmallapuram" in Artibus Asiae Vol. 47 (Zürich 1987), pp. 109–130 at pp. 113–114.)

… or necessarily contemporary with the original buildings in which they’re found. All of this means that three rough groups of approach have been taken to trying to date the complex. The first is to put the different styles of building into a sequence, either monoliths through caves to free-standing or caves to monoliths to free-standing, and then assign them to rulers who fall into that sequence.³ The second is to ignore stylistic issues and follow the lead of the inscriptions, after all a source which was literally designed to tell onlookers about something. The earliest inscriptions that can be dated were put into the shore complex by the Pallava ruler Narasiṁhavarman II Rājasiṁha (ruled around AD 700-728), and because of that and other Pallava rulers attested in inscriptions everyone is agreed that the site is in general a Pallava creation, but what is then done with that varies: one hard-line approach says that since there is no explicit evidence for any ruler before Narasiṁhavarman II having put any of this up, it’s all his, whereas another says that he actually built only the last principal bits.⁴ That is heading for the third approach, which is to try to fit the epigraphical data into a longer sequence, but even one of those writers favours a singular patron figure to have done most of it.⁵ Most people who just want to give a rough number for public-facing purposes, like UNESCO or the Archaeological Survey of India, seem to opt for early seventh-century, but mainly just because of the feeling that a site as big as this must have taken some time to build, so its chronology should be extended backwards from Narasiṁhavarman II. But about the only thing everyone is agreed upon is that he built the Shore Temple as it now stands, which is like this.

The Shore Temple, from closer up this time. You can see at ground level in thje middle of the photo what is left of a ring of instances of Nandi, the bull which Śiva rides, that once encircled the complex

Entrance to the smaller building of the Shore Temple, with the larger in the background

But even this is complicated, because digging in the late 1980s revealed that these buildings originally sat next to a ceremonial tank, which is now visible.

Here is the tank, now sat outside the ring of Nandis

And this opens up another way to sequence the site, from one where many divinities were worshipped and a wide variety of religious stories celebrated – for example the figure in the tank has been read as Varāha lifting the earth from the primordial ocean, to be exposed as the tank was drained – to one which was basically focused on the deity Śiva. The Shore Temple buildings each contain a shrine to him, but are joined by an enclosure you can’t now get into, and it apparently houses a statue of Viṣṇu reclining on a seven-headed serpent, the Anantasayi. That iconography is also seen elsewhere on the site, so that one scholar has suggested that the Shore Temple was a late addition to something already here in the early eighth century.⁶

Barred access to the junction space between the two buildings

And of course, we now have reason to suspect there was perhaps even a small town and certainly other temples out to what is now sea — which may be why Marco Polo, among many other subsequent travellers, referred to this place as the Seven Pagodas when there aren’t obviously seven sea-visible buildings here any more – so something at this site before the Shore Temple doesn’t seem impossible either.⁷ But how much before? And that takes us into the inshore complex, not because that’ll solve the problem (sorry!) but because it adds other factors that need to be considered.

Farewell from the Shore Temple, says this friendly lion with a hidden message I cannot translate for you

The main question that arises here is, how much of it is finished? The monolithic temples all look pretty much complete to the outside eye, and in one case I think everyone agrees that it is, that being the Ganesha Ratha, which stands by itself on the way from the shore.

East end of the Ganesha Ratha, giving you a sense of the medium. Those pillars, for example, aren’t holding anything up, they’re just stone that hasn’t been carved off

Entry side, I think north face, of the Ganesha Ratha. Complete it may be, but its name is probably misleading, as the association with the god Ganesh seems to be later, though to be honest I don’t know how we know

The others also seem outwardly finished, and sit together in a row of four plus an outlier, and some accompanying animals.

Four of the so-called Pāṇḍava rathas (also complicated!) in a row, viewed from I think the north

The Pāṇḍava rathas seen from the other end, including the outlier and the rather fine elephant

Animal pillar bases in the Bhima Ratha

West face of the Bhima Ratha

The Dharmaraja Ratha, seen from the north-east corner

Deities carved into the side of the Dharmaraja Ratha

Nandi, by the Draupadi Ratha, looking pretty relaxed, as he has done for 1400 years or so

Your humble author being friends with a stone elehant, alongside the Nakula Sahadeva Ratha, the one that sits out of the row

You may be able to tell that I love these buildings, and they have themselves occasioned controversy and debate, because they appear to be stone imitations of wooden forms, right down to sculpted replication of jointed roof-beams…

Close-up of corner sculpting on the Dharmaraja Ratha

… and canopies.

The Draupadi and Dharmaraja Rathas, former looking all but thatched

But forms of what? They have been called the Pāṇḍaya rathas for a long time, the first word referring to five legendary brothers who contended with the gods (because there are five of the buildings you see) and the second to temple chariots used for the transportation of divine images in ceremony, which they do look a bit like. But, as one scholar has sharply observed, they don’t have wheels.⁸

It is, admittedly (and he does admit it) possible that the carved animals along the bases of the rathas‘ plinths are meant to be bearers, but that still wouldn’t obviously make these imitation carts, since that isn’t really how carts work

So it may instead be that the carts themselves are imitations of wooden temple forms which were also being imitated here but in carved stone. And that is something that is also seen at a site called Thirucirappali, where an inscription proudly proclaims, or has at least been read as proclaiming, that the technique of excavating buildings as a single piece from the living rock was invented by an earlier Pallava, Mahendravarman I (ruled around AD 571-630), which is how Marilyn Hirsh winds up arguing that he built the inshore complex at Mahabalipuram as well.⁹ But of course, it’s not all built in that technique!

Frontage of the Trimurti Maṇḍapa, which as you can see is more dug into the rock than out of it

Now, admittedly, the main difference here is that the surrounding rock just hasn’t been cleared away; the temple still occupies most of its inside. I’ve read in several places that this is considered unfinished, and maybe that is why, since otherwise it clearly has been operational as a shrine.

One of the entries and its decoration

Entry and interior of one of the shrines

But not all of its slots for statues are full, and there’s a sculptural program around the side of the rock which might be incomplete…

Animal scuplture around the corner, along with a weary visitor

… or might just never have been intended to be more than a resting place blessed by animals beloved of the gods, who knows? This is not the only one of these buildings where I, myself, didn’t realise it was supposed to be unfinished when I saw it.

Front elevation of the Mahishasuramardini Maṇḍapa, which you have to admit looks as if someone meant it to look this way

Pillars and inner carvings

And there are some empty positions in this one, but there’s also these frankly amazing sculptures.

The Mahishamardini panel, depicting the god Durga fighting the buffalo-headed demon Mahishasura

The Anantasayi panel, showing Viṣṇu reclining on the seven hooded serpent Adisesha, that same image apparently sculpted in the bit between the buildings at the Shore Temple

So whether or not this is ‘finished’ seems to hang on whether one expects the whole interior to have been hollowed out, as if it were a normal building, or whether it was in fact meant to be viewed and used from outside, and I haven’t found anyone who thinks they know. But it’s a bit clearer in other cases that not all the intended work got done, and that’s really interesting because it shows you how it was done.

Frontage of the Krishna Maṇḍapa, complete with rocks never removed from the roof

Animal column bases along the frontage

Part-excavated space inside, which I can now sort of explain because of a visit to a Welsh slate mine museum, bizarrely: the chequer pattern in the wall at left and the one at the back is caused by drilling, first inwards horizontally and then somehow breaking into the surface vertically, till it could be unseated in large square lumps and moved on out. Here they may have taken out all they meant to but the wall surface was never finished off

Similarly unfinished cutting in the back of one of the shrine spaces

And again, this building is not complete in and of itself, because there is also a massive exterior sculptural display that goes with it that has to be seen to be believed.

Sculptural relief showing the Descent of the Ganges, with the frontage just shown visible at far left

Central division of the relief, occupied by serpents

The left register, showing an ascetic entreating Śiva to bring the Ganges to Earth, surrounded by other figures

Another animal in the right register, a cat apparently doing yoga, perhaps only possible where gods also tread

So that’s really quite the thing. Now, don’t get me wrong; I can wonder about some individual buildings, but there are more obvious signs that the site as a whole was not fully realised according to plan.

This carving was abandoned before it even really started…

… these steps don’t go anywhere and never did…

… and this got some way, but they apparently started at the top and worked down, then stopped

And the story this seems to tell is of a really massive program that encompassed the whole site, even if not necessarily all at once—though if you are either Hirsh or Nagaswamy, then yes, all at once, pretty much—which had its patronage quite suddenly withdrawn.¹⁰ (I can’t help but analogise to Domesday Book, which gives us exactly the same problem of a programme stopped before we can be quite sure what it was for.) The obvious explanation for that might be a change of reign, but it might also have been a radical reprioritisation towards war; again, we don’t know, and many different theories are sustainable. But it is at least arguable that despite the myriad buildings and complexes here, it was never actually seen into use as intended, and that subsequent shrines, ceremonies, additions and so on were attempts to patch into the indubitable sacrality of a program nevertheless gone unrealised.

Frontage of the Rāmānuja Maṇḍapa, viewed from the side

It’s a secret that maybe only the sculptures, or the divinities they represent, will ever know.

Sculpture around the entry to the Olakkanesvara Gopura

I reckon these two know more than they’re saying, for a start

Local residents on the Olakkanesvara Gopura who probably also have their secrets

Occupant of the Shore Temple of whom like suspicions might be maintained

The weird thing is, it seems almost as if the problem was catching, because – unless those who think the Shore Temple is the last major bit are right, and I’ve got no opinion, that’s for sure – it doesn’t seem as if anyone else was ever able to complete anything here either. In fact, it was a while before anyone else tried; apparently there is later sculpture here, including stuff from the tenth century, but I couldn’t identify any of it from signage and the main event is this, a gopura of the fourteenth-century Vijayanagara period that sits on top of one of the cave rocks…¹¹

The Rayar Gopuram, as one approaches it

… but which never seems to have got its roof, assuming that one was intended.

West half of the structure from closer up

Central roof supports, sans roof

I mean, perhaps by then an unfinished building was considered the appropriate thing to add to this site so as to participate in it; perhaps the local stories and mythologies had encompassed that aspect of it. Of course, we’ve no idea. But it does make me wonder.

Another unfinished structure, on the next hill inland from the Rayar Gopura, about which I could and can still learn nothing

Mainly, what it makes me think is that unlike the busy over-building that I was trying to unpick at Rome and Istanbul, or the continuous maintenance and updating that was visible at Nánshā in China, another shore temple of a different sort, this site has spent most of its historical life partway abandoned, available, in use, but never at the scale or for the purpose its designers may have had in mind. And it’s hard to see how we’ll ever know what that was, either. But it’s still amazing to see and I hope very much to be able to go back and look again some day.

View from atop the Mahishamardini Maṇḍapa

---

1. We’ve already mentioned the problems asserting Middle Ages that apply in India, but there is at least a scholarly defence for the application of late Antiquity to the Indian Ocean world, Rebecca Darley, “Byzantine Gold Coins and Peninsular India’s Late Antiquity” in Sven Günther, Li Qiang, Lin Ying & Claudia Sode (edd.), From Constantinople to Chang’an: Byzantine Gold Coins in the World of Late Antiquity, Supplements to the Journal of Ancient Civilizations 8 (Changchun 2021), pp. 135–173. Rebecca has also helped a lot with this post, so all thanks to her.

2. “Tsunami Uncovers Indian Shrines” in Science Vol. 308 no. 5720 (New York City NY 2005), p. 350; Ekatha Ann John, “Millennium-old ‘sunken town’ found off Tamil Nadu” in Times of India (New Delhi 19th March 2016), sec. India News, online here.

3. The former sequence seems to be the approach of K. R. Srinivasan, Cave Temples of the Pallavas (New Delhi 1964), which I haven’t seen but is cited to this effect by Marilyn Hirsh, “Mahendravarman I Pallava: Artist and Patron of Māmallapuram” in Artibus Asiae Vol. 48 (Zürich 1987), pp. 109–130 at p. 128. Raju Kalidos, “Stone Cars and Rathamaṇḍapas” in East and West Vol. 34 (Rome 1984), pp. 153–173, goes for the other chronology.

4. The case for Narasiṁhavarman II is made by R. Nagaswamy, Mahabalipuram (Mamallapuram), 3rd ed. (New Delhi 2010).

5. Hirsh, “Mahendravarman I”.

6. John R. Marr, “Note on the new excavations at the Shore Temple, Mahabalipuram” in Bulletin of the School of Oriental and African Studies Vol. 54 (London 1991), p. 574.

7. Aniruddh S. Gaur and Kamlesh H. Vora, “Maritime Archaeological Studies in India” in Alexis Katsambis, Ben Ford and Donny L. Hamilton (edd.), The Oxford Handbook of Maritime Archaeology (Oxford 2011), pp. 513–534 at pp. 525-526.

8. Kalidos, “Stone Cars and Rathamaṇḍapas”, pp. 154-160.

9. Hirsh, “Mahendravarman I”, pp. 116-127.

10. See nn. 4 and 5 above.

11. The chronology here is from the site signage.