Skip to main content

The Indian Temple: Production, Place, Patronage

This project was a collaboration with the School of Oriental and African Studies (SOAS) and The British Museum to look at how temples were patronised and constructed and the place they occupied in a medieval Indian polity.

The project began in 2006, with a grant of £632,186 f.e.c. for three years from the Arts and Humanities Research Council of the United Kingdom.

Project background

Temples dominated the landscape of India between the seventh and thirteenth centuries. Protected by kings and widely supported by endowments and other gifts, temples enjoyed ascendancy as centres of religious life, socio-economic power and artistic production.

Although much research has been carried out on temple architecture since the late nineteenth century, important questions remain about how temples were patronised and constructed and the place they occupied in a medieval Indian polity. Three general themes, each with a corresponding question, provided a framework for our research:

  • Production: How were temples designed and built? Led by Adam Hardy, Cardiff University.
  • Place: What was the social and political role of medieval Indian temples? Led by Michael Willis, The British Museum.
  • Patronage: What was the king’s role as a patron of architecture and Sanskrit letters? Led by Daud Ali, The School of Oriental and African Studies.

The ultimate aim was to achieve an integrated understanding of these questions, establishing a new model of interdisciplinary research in the field.

Seminar: the Medieval and its contexts

An international seminar was held at the Royal Asiatic Society, London and SOAS, University of London, 2-3 July, 2009. The purpose of the seminar was to assess the nature and characterisation of the 'medieval' by contrasting and comparing the medieval religious architectures of India and the west and by examining archaeological, theoretical and historiographical developments in the study of the medieval in India and Europe. Issues included the nature of state formation, the relationship of agrarian production and trade to the socio-religious order and the use of surpluses by religious institutions for major building projects.

Speakers and discussants

  • Daud Ali (SOAS)
  • Prof. Hans T. Bakker (Groningen)
  • Nora Berend (Cambridge)
  • Anne Casile (Cardiff University)
  • Whitney Cox (SOAS)
  • Peter Draper (Birkbeck College)
  • Paul Dundas (Edinburgh University)
  • Adam Hardy (Cardiff University)
  • Nigel Hiscock (Oxford Brookes)
  • Kathleen Morrison (University of Chicago)
  • S. Reynolds (University College, London)
  • Michael Willis (British Museum).

The temple site at Bhojpur

In addressing these broad themes, one temple site, paradigmatic of medieval Indian in general, was taken as the starting point for the project. This is the site of Bhojpur, near Bhopal in central India, where a gigantic temple was left unfinished in the mid-eleventh century. Around the temple are quarries and unfinished architectural parts, along with architectural drawings engraved on the rocks, a unique survival providing insights into the processes of design and construction.

The temple is associated with King Bhoja of the Paramara dynasty, a renowned polymath. Bhoja is the subject of many stories and legends in nearly every Indian language, and is known as one of the greatest kings of medieval India. Over thirty literary works are ascribed to him, including the Samaranganasutradhara, a famous but barely understood architectural treatise, which formed a central part of the study.

The team

Our two main collaborators in India were Dr O. P. Mishra (Department of Archaeology, Madhya Pradesh) and Dr Vishwa Mohan Jha (Delhi University).

Two Research Assistants took part in the project. Mattia Salvini, based at SOAS, worked on the Samaranganasutradhara. Anne Casile, based at Cardiff University but working with Michael Willis at the British Museum, created a database of archaeological remains and fragments of temples in three districts of Madhya Pradesh.

The project doctoral student, Doria Tichit, completed her PhD on the Udayeshvara, the exquisite Paramara temple at Udayapur.

Amita Kanekar did extensive survey and drawing work, and there were important contributions in this respect by Ananya Gandtora, Sonal Mithal Modi, and Yashaswini Sharma.

An international seminar exploring the nature of 'the medieval' in a comparative context was held in 2009.


The starting point of the project was the unfinished eleventh-century temple at Bhojpur attributed to King Bhoja. Bhojpur is unique in its preservation of quarries, an earthen ramp, and many unfinished architectural parts, of which we have made measured drawings. Their size shows the vast scale that the finished monument would have attained. Most interesting of all are the numerous engraved architectural drawings on the rocks, which have been recorded (in measured line drawings) and interpreted for the first time.

In parallel, we have analysed another work ascribed to Bhoja, the Samaranganasutradhara, a famous but barely understood architectural treatise. For the first time substantial passages have been translated, by Mattia Salvini, in a way that is architecturally meaningful. This study has culminated in Adam Hardy's Theory and Practice of Temple Architecture in Medieval India.

Both the engraved drawings and the treatise throw new light on temple design. But neither drawings nor text make sense without a thorough understanding of the buildings themselves, including principles of composition, geometry and measurement, and construction. The project has enabled us to carry out surveys of a wide range of temples in central and western India and the Deccan. Constructional techniques have been examined in ruined temples, and in the normally unseen spaces in roofs and towers.

Latina temples

The Latina form, with its curved spire (shikhara), developed from multi-tiered shrines known from the Gupta period (fourth-fifth centuries AD) onwards, with pavilions crowned by amalakas (rounded, ribbed element) at the corners of their false storeys. It was predominant type of Nagara (north Indian) temple between the seventh century and the tenth, when it was overshadowed by its offspring, the multi-spired Shekhari (or Anekandaka) mode of Nagara.

Some obvious questions arise about design and construction, and we have been exploring these through measured surveys. Do the plans follow grids? Is there a module for plan and elevation, or are there several? Does the geometry of the gavakshas (arched dormer motifs) relate to an all-encompassing system? What is the nature of the curvature of the shikhara, and of its constituent segments, and how was this achieved? What are the rules for successively diminishing the stages of the superstructure?

Latina temples have been measured and studied at the following sites:

  • Madhya Pradesh: Amrol, Bateshara, Gwalior (Gujari Museum), Nareshara and Terahi.
  • Gujarat: Roda, Shamalaji and Modhera.
  • Rajasthan: Osian.
  • Karnataka: Pattadakal.
  • Andhra Pradesh: Alampur.

Shekhari temples

The composite Shekhari (or Anekandaka) mode was the most important form of Nagara temple from the tenth century onwards in central and western India. It is conceived as a constellation of interpenetrating shrines burgeoning out from the single spired form of the Latina temple. Predominant among the embedded shrine forms is the form of the Latina itself, self replicating. The kuta-stambha – a miniature shrine or shikhara crowing a pillar – is a basic element that needs to be recognised in order to grasp the thoroughly composite or multi-aedicular nature of this kind of temple.

As Shekhari temples are a continuation of the same Nagara tradition, the questions relating to the design and construction of Latina temples are also relevant here. Further questions arise from the complexity of their geometry in three dimensions. Adam Hardy proposed geometrical principles in an earlier study based on observation rather than measurement (‘Sekhari Temples’, in Artibus Asiae 62, No. 1, 2002, pp. 81-137). Are these confirmed by measurement? How closely do the actual buildings conform to the ‘ideal’ geometrical figures? Our research is based on measured surveys in Gujarat, at Asoda, Modhera, Sejakpur, Sunak and Sander.

Several chapters of the Samaranganasutradhara deal with Nagara temples, unfolding elaborate typologies from unitary Latina forms to hyper-proliferated Shekhari ones. We are working on translating these, both into English and into drawings. Later texts from western India treat the design of Shekhari temples systematically, and the tradition is perpetuated today in Gujarat by the Sompura caste of traditional architects. We are studying two modern Sompura texts, the Shilaparatnakara from the 1930s, and a 1960s Gujarati commentary on the Kshirarnava. A planned future study will combine measured analyses of medieval temples, an illustrated translation of the Aparajita (c. thirteenth century), studies of later texts, and a critical study of the working practices of present-day traditional practitioners.

Bhumija temples

The Bhumija mode appeared in the eleventh century as an alternative composite form to the Shekhari, and is also clearly an extrapolation from the earlier Latina. It is recognisable by its radiating chains of kuta-stabmhas (spires on pillars) between the bhadras (central projections), and can be built on an orthogonal or a stellate (rotated square) plan. Bhumija temples are especially relevant to this project as this was the temple form favoured by the Paramara dynasty in Malwa. Here and elsewhere there seem, on the basis of style, to have been guilds of temple builders specialised in this form. Bhoja’s unfinished temple at Bhojpur, and the surrounding drawings and fragments, belong to this tradition. It is likely that the temple was intended to be a Bhumija structure of vast proportions, of which the giant cube visible today would have been merely the inner sanctum.

Aspects being explored include the geometry of orthogonal and stellate plans and, once again, the questions of diminution and curvature in the superstructure. Here we are lucky to be able to compare measurements of actual temples directly with drawings and textual prescriptions. At Bhojpur a large drawing survives on the stone platform directly in front of the sanctum. As might be expected of a text attributed to the same patron, Chapter 64 of the Samaranganasutradhara, the chapter concerned with Bhumija temples, contains vital insights into their theoretical typology and proportions.


The study of the shrine itself (mula-prasada, vimana) cannot be seen in isolation from the hall (mandapa) to which it is generally attached – its overall form with its roof shape, its pillars, its ceiling or ceilings. In northern traditions the mandapa usually follows the tiered Phamsana form (also a shrine form) or, from around the tenth century, the composite Samvarana form. In its proliferating evolution the Samvarana mirrors the development of the Shekhari shrine form, with which it is often associated, and a parallel pattern can be observed in the blossoming of multi-lobed corbelled ceilings.

A typical madapa structure is open at the sides, with a stone seat around the perimeter, pillars and beams, an awning-like eave slab (chhadya), and corbelled ceilings surmounted by a corbelled pyramidal roof. Indian temple studies have taken an armchair and arms-length course, and it is characteristic that not a single constructional cross-section of this kind of madapa has ever been published. We shall soon remedy this.

Dravida temples

It might seem surprising that a project beginning its investigations at Bhojpur in Malwa (in present day Madhya Pradesh) should concern itself with Dravida temples from south India, but there are several reasons why it needs to. The architects of the Bhumija temples in Malwa were well aware of Dravida architecture. Even this far north, Bhumija temples, though broadly Nagara in character, show stylistic affinities with contemporary Dravida traditions in the southern Deccan (Karnataka, Andhra).

Samaranganasutradharaas a Dravidakarma kuta. Two of the engraved line drawings at Bhojpur depict a form of mandapa unknown in any surviving building, with a roof composed of these Dravidakarma components. Fragments of this kind of kuta are found at Bhojpur and among the ruins of another huge Paramara temple, the Bijamandal at Vidisha. The Samaranganasutradhara contains a substantial section devoted to Dravida temples, which originated in Tamil Nadu, far to the south, as shown in a recent article by Adam Hardy.

Our measured studies have concentrated on Karnata Dravida temples (often classified as Vesara) of the eleventh and twelfth centuries. As with Nagara temples, though without the issue of curvature, an important question is how diminution was achieved in the stages of the superstructure.

We have done measured surveys of temples built under the Later Chalukyas and Hoysalas at the following sites in Karnataka:

  • Aralguppe
  • Badami
  • Balligave
  • Belavadi
  • Banashankari
  • Chaudanpur
  • Damba
  • Halebid
  • Hangal
  • Ittagi
  • Lakkundi
  • Sudi
  • Turuvekere.

A Karnata Dravida temple in ‘Bhumija territory’ has also been surveyed - the Ishvara or Ayeshvara at Sinnar (Maharashtra).

This research has proved invaluable in one unanticipated and direct application: a commission for PRASADA to design a new ‘Hoysala’ temple for the Sri Kalyana Venkateshwara Swamy Temple Trust.

Placing the Gods

This was a short design project run as one of the Welsh School of Architecture’s Vertical Studio options in 2008. It involves a group of first- and second-year Architecture students. The Vertical Studios are intended to feed some of the School’s research into its teaching. A total of 11 students took part in this option.

Project brief

An objective of ‘The Indian Temple’ project is to catalogue a section of The British Museum's magnificent collection of Indian sculpture (c. 2nd century BC to 13th century AD), and to explore how this collection might be more meaningfully and imaginatively displayed. Its contents were originally conceived, made and venerated as parts of subtle iconographic programmes, which in turn formed just an aspect of great and complex sacred monuments. Once integral to an architectural whole, the sculptures now sit as specimens in cases or on pedestals.

The vertical studio group visited The British Museum and were given special access to its reserve collection. Each member studies pieces, while the entire group considered the display of the collection as a whole.


Project team and collaborators

Project leads and main collaborators

Professor Adam Hardy

Professor Adam Hardy

Emeritus Professor

+44 (0)29 2087 5982
A profile image of Michael Willis

Michael Willis

Curator, South Asia, The British Museum

A profile image of Dr Daud Ali

Dr Daud Ali

The School of Oriental and African Studies (SOAS), University of London

Project team

  • Fiona Buckee
  • Anne Casile
  • Neeta Das
  • Meera I. Dass
  • Ananya Gandtora
  • Vishwa Mohan Jha
  • Amita Kanekar
  • Ajay Khare
  • O. P. Mishra
  • Sonal Mithal
  • Pranali Parikh
  • Mattia Salvini
  • Yashaswini Sharma
  • Doria Tichit
  • Archana Verma

Other collaborators

Thanks are due to the Department of Archaeology, Madhya Pradesh and the Department of Archaeology, Uttar Pradesh, for their collaboration and support. We are grateful to the Archaeological Survey of India for their cooperation and for granting permissions, especially to the following circles: Aurangabad, Bangalore, Bhopal, Dharwad, Jaipur, Lucknow, Mumbai, Vadodara.

The project is grateful to the following institutions, organisations and people for their help with fieldwork:

  • School of Planning and Architecture (SPA), Bhopal
  • School of Architecture, Birla Institute of Technology, Ranchi
  • School of Architecture, Lucknow
  • DRONAH (Development and Research Organistaion for Nature, Arts and Heritage)