South Asjan Studies 12 1996

Architecture in Central India under the Kacchapaghata

Rulers
MICHAEL D. WILLIS

Introduction therefore, is to build on Krishna Deva's foundation.

This will be done through a discussion of the better-
Numerous temples are preserved in the territory known monuments as well as several temples which
between the Gangetic plain and the Narmada river. are published here for the first time.
In ancient times, this vast area (conforming A few preliminary comments are necessary
approximately to Madhya Pradesh) consisted of regarding Sanskrit terms and place names. Technical
several regions with differing cultural and terms in Sanskrit are transcribed according to the
architectural traditions. The northwest portion, currently accepted system. Modern names of temples
round the hill-fort of Gwalior, was known as and locations are not Sanskritized and some attempt
Gopak~etra from at least the 11th century (Sinha, has been made to account for current pronunciation
1980; Willis, 1988). It was in this region that the in Hindi. Where possible, place names are spelt
Kacchapaghiita dynasty flourished as the great according to the maps published by the Survey of
dominion of the Gurjara Pratiharas was parcelled India and documented in Corpus Topographicum
into a number of independent kingdoms. The most lndiae Antiquae (Part I, Epigraphical Find Spots)
powerful neighbours of the Kacchapaghatas were by R. Stroobandt (Gent , 1974).
the Candellas (to the east in ancient Jejakadesa),
the Paramiiras (to the south in ancient Maiava) and
the Cahamiii:tas (to the west in ancient Virata and Kacchapaghata Dynasty
other parts of Rajasthan). While the Pratihiiras
struggled to retain Kannauj and the Gangetic plain, The Kacchapaghata dynasty came to prominence in
these and other dynasties vied for political the 10th century after they were able to capture
supremacy. In the competitive environment that the fortress of Gwalior from the Gurjara Pratihiiras.
prevailed, architecture enjoyed an extraordinary The Sas Bahu temple inscription, a record which
efflorescence and temples of unprecedented scale will be mentioned frequently in the following pages,
and complexity were constructed. recounts that Vajradiiman (c. A.O. 975-1000) "put
The architectural achievements of the Candellas down the rising power of the ruler of Gadhinagara
are well known, no doubt because their buildings [Kannauj] and his proclamation drum ...resounded on
are in an excellent state of preservation (see Deva, the fort of Gopadri" ·(Trivedi, 1979-91, p. 542, v.
1959, 1987). By comparison, the temples of the 6). The Kacchapaghatas subsequently ruled from the
Kacchapaghiitas have been ignored. Setting aside the twin centres of Sihoniya (ancient Sirilhapaniya) and
desultory treatment in survey books (for example, Gwalior (ancient Gopadri, Gopagiri); collateral
Brown, 1944 [?]; Rowland, 1956, p. 169), the only branches of the family had their seats at Dubkund
examination of later temple architecture in the (ancient Oobha) and Narwar (ancient Nalapura).
Gwalior region has been provided by Deva (1963- We can show the sequence and distribution of the
64, 1969). Of necessity, these contributions provide known Kacchapaghata princes as follows (based on
a first introduction to the subject. Our purpose here, Dvivedi 1980 with slight modifications):
14 Michael D. Willis

Kacchapaghata Dynasty

Simhapaniya/Gopadri l;>obha Nalapura

Lak~ma1.1a (c. 950-75)
Vajradaman (c. 975-1000)
Mailgalaraja (c. 1000-1015) Yuvaraja (c. 1000)
Kirtiraja ( c. 1015-35) Arjuna (c. 1015-35)
Miiladeva ( c. 1035-55) Abhimanyu (c. 1035-45)
Devapala (c. 1055-85) Vijayapfila (c. 1045-70)
Padmapala (c. 1085-90) Vikramasirilha (c. 1075-1100) Gaganasirilha (c. 1075-90)
Mahipala (c. 1090-1105) Saradasirilha (c. 1090-1105)
Ratnapala (c. 1105-30) Virasirilha (c. 1105-25)
Ajayapala ( c. 1192-94) Tejaskara1.1a {?]
Sulak~a1.1apala (c. 1196)

An examination of the accompanying map shows Dubkund Kacchapaghatas has been known since the
that each branch controlled a geographically distinct site was explored in the 19th century; it is dated VS
area (Fig. 1). The Gwalior line, apparently the most 1145 bhadrapada sudi 3/A.D. 21 August 1088
powerful and certainly the most important in terms (Trivedi, 1979-91, 3: p. 530). The inscription, now
of documented architectural activity, ruled the large in the Archaeological Museum, Gwalior, was found
tract bounded on the north and west by the river in the ruins of a large Jaina temple and records that
Chambal (ancient Carma1.1vati).The Narwar branch the building was constructed by an individual named
apparently controlled the area south of the river Dahac;la and several other eminent Jainas. The
Parbati (ancient Parvati). This part of the construction was also supported by King
Kacchapaghiita domain perhaps went as far south Vikramasirilha who further made revenue donations
as Surwaya, beyond which lay the territories of in the temple's favour. The building consists of a
some later Pratihara princes (Dvivedi, VS 2004, square cloister (24.5 m each side) once edged with
nos. 627-33). Still further south was the Paramara shrines carrying roofs of pyramidal form
kingdom. The river Betwa (ancient Vetravati) (pharilsana'). The doorjambs are richly carved as are
marked the edge of Candella territory in the the pedestals and canopies. The complex was in an
east; Candella inscriptions are found at Jhansi, advanced state of ruination even in the 19th century
Deogarh and Mohangarh (the latter unedited but (for illustrations see Luard, 1908, plates 159-62).
discussed in Willis, 1993, p. 55). The Dubkund The plan of the Dubkund temple recalls monuments
Kacchapaghatas controlled lands up to the Chambal at Mount Abu and elsewhere in western India. Such
in the west. influences are also suggested by the remains at
Gwalior, as we shall see below. Another inscription
at Dubkund was found on the base of a Jaina image
Narwar and Dubkund (Dvivedi, VS 2004, no. 58). It is dated VS 1152
vaisakha sudi 5 and mentions a teacher by the name
The temples at Narwar (ancient Nalapura) have of Sri Devasena.
been destroyed, with fragments incorporated into The support of Jainism by the Kacchapaghatas ·
later buildings. Kacchapaghata monuments in the raises some hitherto unaddressed questions about
immediate neighbourhood have· been surveyed this faith in connection with temple patronage.
(Patil, 1952), but have not been subject to systematic Jainism has been defined by lndologists as a
study. At Dubkund, there are Kacchapaghata minority religion ; its mythology, theology and ritual,
shrines but these have been reduced to foundations aside from resonances With archaic forms of
and base mouldings. The ruins were uncovered in Buddhism, are often viewed as things of strictly
the early 1980s by the Department of Archaeology special or parochial interest. While it is true that
and Museums, Madhya Pradesh, but a full report the Jainas have been remarkably careful in
of the excavation does not seem to have been preserving their traditions and, like most
published as yet. A long inscription mentioning the communities, concerned mainly with their own
 fu ro numbered si/cs
I DATESAR
2 PAROLI
3 MITAOLJ
4 DHAINSORA
KOTWAL
6 DARAfiET

AMROL•

• DUD KUND

SATANWARA •
I '
• DALLARPUR ,;:-
._
~
' G"P~'AM,U
SIIIVrURI
KOTA
• • • ) ~ • HATIIU.U
SURWAYA
SESAI UUZURG. Sil.OAR
PATIIART •

Ri!.J.lt.PUR
KHELDUAR • RANNOD •
•7TER.l.Hl
SAKARRA • MAIIUA

INDOR •
• •KADW!l.l!A

• MAIIUAN • MARKIIERA

1 0 • TUMAIN UMRI
:1[
IllIIJ•Jl:I
r::1 =='fc:::::::ii°
Kilomclrcs

Fis, .1. Map showing the principal sites in the Kacchapaghat a territories.
16 Michael D. Willis

affairs, the picture of Jainism as a small sealed world especially rich in remains of the period, the most
does not square with the fact that several minor magnificent being a large Siva temple locally
Indian dyasties, notably the Gailgas in south India, designated Kakanmath. The inscription on the Sas
were patrons of significant Jaina monuments. An Babu temple at Gwalior (discussed below) provides
explanation of this can be found in the beliefs and clues about the authorship and date of this building.
practices of the faith (Handiqui, 1968; Babb, 1993). The key verse is as follows (Trivedi, 1979-91, 3: p.
We cannot enter into all the details, but the most 542, V. 11:
illuminating feature is that during elaborate pujas,
J aina worshippers often wear crowns to symbolize adbutai) sirilhapaniyanagare yena karitai)I
their identity with Indra or Indrai:ii. This seems to kirttistarilbha ivaohati prasadal) parvvatipatel)/1
be an ancient practice once shared with Buddhism.
In the town of Simhapaniya he [Kirtiraja]
Setting aside the early Buddhist texts and Ku~ai:ia- constructed a wonderful temple Parvati's Lord,
period images which show the Buddha attended by
Which shines like a column of glory.
Indra · and Brahma, the first known instance of a
devotee actually dressing himself as Indra is Har~a Sihoniya can be equated with Siril.hapaniya and
Vardhana (r. AD. 606-647). This is documented by the Kakanmath with the temple of Parvati's Lord
Xuan Zang in his description of a royal festival at (i.e. Siva). An earlier verse, which we need not
Kannauj (Watters, 1904). It seems unlikely that quote for the present purposes, describes Kirtiraja's
Har~a's was an isolated case. The worshipping of a exploits against the Paramaras, indicating that the
Buddha or a Jina would certainly have shown the word yena (by him) refers to none but Kirtiraja. As
devotional inclinations of a monarch, but Xuan indicated above, this king ruled c. 1015-1035. A date
Zang's account shows that the accompanying in the first part of the 11th century for the
pageantry and processions were intended to publicly Kakanmath is confirmed by the general similarity
display the king as Indra. Now it is commonplace of the building to the Visvanatha temple at
in Hindu mythology that Indra is the king of the Khajuraho. The Kakanmath stands on a broad
gods; he occupied this position in Vedic mythology terrace and was originally surrounded by at least
and is still regarded as such today. As a four subsidiary shrines. The temple-proper consisted
consequence, the ritul actions of an earthly king of a sanctum surrounded by an ambulatory with
clothed as Indra \\'ere not so much intended to balconied transepts; in front of this was a vestibule
associate the king and the image under worship, but (antarala) and a closed hall (gu,;Jhamal),;/apa)with
to reinforce the king's identity with Indra. By lateral transepts and a porch approached by stairs.
imitating Indra as an admirer, supporter and helper The massive lions that flanked the steps have been
of the Jina, a king affirmed his privilege to be an removed to the Archaeological Museum, Gwalior,
Indra among men (narendra). This helps explain and now stand on either side of the museum's
why rulers could pay their respects to a tirtharilkara entrance building. Many single images have also
without taking on the faith of Jainism, an impossible been taken to the museum or the Archaeological
prospect in any event given the inherent divergence Survey stores on Gwalior fort. The socle (pitha) of
of Jaina doctrine and the responsibilities of kingship. the Kakanmath is elegantly moulded and punctuated
The support of Jainism by the Kacchapaghatas was by niched figures of deities (Fig. 2). The lofty plinth
thus neither an indication of their acceptance of the used at Khajuraho has been avoided. The sanctum
articles of Jainism nor an indication of their is girded by the standard podium mouldings, pierced
tolerance in the modern sense. Rather, it was a on the north side by a spouted channel (pral)lila)
means by which they could puolicly invoke their used to draw libations away from the liilga in the
royal status. cella.
The outer walls, balconies and facing stone of
the spire have all fallen· away, probably as a result
Gwalior and Sihoniya of an earthquake. The only preserved portion of the
elevation is that of the sanctum itself. This shows a
The buildings discussed thus far pale beside the single row of large images in niches, one on each
major temples at Sihoniya and Gwaiior. Sihoniya is offset. The images are framed by pilasters and
 Kacchapaghata Temples 17

miniature arches spewing from th~ mouths of

grotesques (makaratoraQa). The intervening
projections and recesses are filled with nymphs and
leogryphs ( vyala). Figural friezes run across the top
.of the wall. The cardinal projections (bhadra) display
larger niches, now empty (Fig. 3). Following 9th-
century conventions, these niches take the form of
miniature doors with ribbed awnings and phiirilsanii-
style pediments. The sanctum doorway has seven
ornate jambs (sakha"), one consisting of a large row
of deities between two bands of couples. The sill
( udambara) carries elephants and recumbent lions,
the latter outstanding examples of llth-century

Fig. 3. Sihoniyii, Siva temple, cardinal niche, first quarter

of the 11th century. Photograph courtesy of Donald M
Stadtner.

sculpture and its tight linear style (Fig. 4). The spire
survives to its full height (approximately 37 m) but
only the rough masonry core and part of the
serrated crown (iimalasaraka) have survived. Fallen
aedicules (ktifa) give some hint of the spire's original
appearance. As indicated by the Siva temple at
Kadwiiha (in the so-called Morayat group) , these
aedicules were probably used at the base of the
main spire (miilasrliga). The vestibule and the hall
of the Kakahmath have been shorn of their outer
walls, revealing clusters of pillars simply ornamented
Fig. 2. Sihoniyii, Siva temple, outer wall of vestibule, first on the upper quarter and crowned by plain roll-
quarter of the 11th century. brackets. The central ceiling is lost, but peripheral
18 Michael D. Willis

Fig. 4. Sihoniyii. Siva temple . door sill. first quarter of the 11th ce ntur y.

ce ilings show d esigns of cusped coffers. Th e hall rose at Sihoniya. Many hav e be e n co llec ted in a mod ern
to a height of thr ee st oreys and was crowned by a Jaina temple and dharma sa la to th e south of the
bell-finial (gha(lfa). village , but others were still in th e open in th e 1980s
T he Kakanmath is not the only building at (Fig. 6). Th e mo st venerated templ e at Sihoniya is
Sihoniya belon ging to the Kacchapa ghata period. that of Hanum a n. Th e s hrin e is a mod ern
Fragmentary sculptur es are scattered throughout th e construction o n a large mound to the north of the
village and architectural member s indicate that ther e village; it rests on old foundations and the re are
were once important templ es th e re. Among the various carved fragment s sca tt ere d abo ut. The
many pieces which could be discussed, perhaps the colossal Hanuman ima ge is quit e impr ess ive and
most poignant indicat or of what has bee n lost is the ma y belong to the end of th e Ka ccha pag hata period.
large aed ic ule (k iifa ) illu strat ed he re in Fig. 5. The temple of Ambika Devi, o n the western edge
Compared to the examples at the Ka kanmath , ju st of the villag e, is also of co nsiderable sanct ity. This.
discussed, this kii/a has a more intri ca tely worked building is se t in a cloistered compound
mc5h-?a\tern (Ja.Ja), indicatin g that it is so mewhat incorporatin g columns, capital s, brackets and ima ges
later in date. The accompanying pillar with festoons of var iou s type s and dates. The o lde st material
and att ach ed figur es is coeval. These device s were belongs to the 9th cen tury. The templ e has been
particularl y popular in Gopaksetra during the lat e heav ily reco nst ruct ed in a haph aza rd ma nner using
11th century, as will be seen below. In addition to a variety of antique p art s. This is also tru e of a
this mat eri al, num ero us Jaina images ca n be found seco nd shr ine in th e compo und directly adjacen t to
 Kacchapaghata Temples 19

, .....
..
.._. -~w-.-•
~-.l

~ ...-~;
~ ) ~

¾
Y'- --- -
,_,i;J. ,
.
ivtnl;.
'•...
w,e-

Fig. 5. Sihoniyii, architectural parts, to the left with Variiha image , 9th century, to the right 11th century.

the Ambika Devi. The entrance door of the second prajabharttra tena k$ititilakabhr1tena {bhajvanam
shrine , however , is reasonably well preserved (Fig . harerddharmmajnena trida$asa[dr$ti}karitam
7). Among the many interesting pieces in this ada/JI
complex, a fragment showing a dancing figure may
be singled out for attention (Fig. 8). Assignable to vadamyasyo[ccaistvamj katham iva gira yasya
the mid -lOth century, this piece was originally part sikhararil samarii(fha/J sirilho mrgam iva
of a temple wall Uangha). The dancer has the [mrgankajstham asitum II
twisting dynamism often associated with later Indian This supporter of his subjects, an ornament of
art, but the sculptor has happily avoided the ugly the earth, who knew the law (and) who was like
distortions and fragmented modelling of the body the gods, commissioned this edifice of Hari.
which is found in many of the female images on
the Lak$ma1:1citemple , Khajuraho . How can I possibly tell you of its height with
Sihoniya is not far from Gwalior and in that words ? The lion who has mounted the spire
great fortress there are two Vai$1.1ava temples which means to devour , it seems, the deer in the moon .
mark the culmination of the Kacchapaghata style .
Popularly known as Sas Bahu ("daughter - in-law and The inscription informs us , however , that
moth er-in-law "), the larger temple preserves long Padmapala died before compl eting the temple (v. 30)
inscriptions in the porch recording that the building and that Mahipala , as soon as crowned, vowed to
was commenced by King Padmapala ( c. 1085-90). finish the building (vv. 69-70). This he did , as well as
One of the key verses in the dedicatory inscription providing numerous endowments , gifts and valuable
is as follows (Trivedi, 1979-91, 3: p. 543, v. 26): treasures . The temples contained several sacred
20 Michael D. Willis

Fig. 7. Sihoniya , Ambika Devi temple , shrine entrance

door , 11th century .

east side has a large two-storeyed porch. This porch

shelters the inscriptions and the main entrance (Fig.
9). With its spiral columns, profusion of jambs,
complex sill (udambara) and over-door (uttaraliga),
it is the superlative example of the mature
Kacchapaghata idiom. The interior is a spacious
Fig. 6. Sihoniya, seated tirtharhkara , 10th century.
twelve-sided ball with an enormous circular ceiling
"images, but the most important was Sri Padanatha of concentric rings, supported by four massive piers.
(i.e. Vi~i:iu) named in honour of king Padmapala. In the corners of the hall are shrines , now empty.
The larger Sas Bahu temple comprises a These represent an internalization of the
sanctum , ve~tibule and closed hall (gu<;ihamaJJ(fapa). paiicayatana or quincunx plan, an arrangement seen
Only the much -restored core of the sanctum earlier in the Gwalior region , most notably in the
remains , the outer walls and superstructure having hall at Paroli , a building that can be placed in the
disappeared. The hall, however, has been preserved early lOth-century prior to the rise of Kacchapaghata
and exhibits grand dimensions and an unusual power. The images on the exterior of the Sas Bahu
design. Rising to a height of 24.4 m in three-storeys, are defaced , but enough has been preserved to show
it is surmounted by an elaborate pyramidal roof the remarkable decorative vitality of surface, with
covered with miniature spires (sariJvaraQa} This is the figures combining tightly modelled forms with
different from Khajuraho where roofs with simple agitated linear movement (Fig. 10). The interior
slabs (phamsana"') were preferred. The north and retains much of its original appearance with stencil -
south sides have broad pillared porticoes while the like ornamentation covering every available surface.
 Kacchapaghata Temples 21

Fig. 8. Sihoniyii, Ambikii Devi temple , dancing figure, Fig. 9. Gwalior fort, larger Siis Bahii temple, entrance
10th century . door, dated by inscription VS 1150/AD 1093-94.

The temple has been subject to extensive and appear in the Sas Bahii may represent influence
somewhat intrusive restorations. Large beams have from western India , where similar forms are found.
been added inside for support and , on the exterior, However, in view of the fact that so many buildings
the mouldings have been extensively rebuilt in many have disappeared we must be extremely cautious
places. In addition, some of the exterior walls have with such observations. Above the mouldings is a
been completely rebuilt and alien pieces (some dating parapet wall with a sloping seat (kak$lisana). On
as early as the 9th century) have been introduced. the seat are dwarf pillars which support a ribbed
The smaller Sas Babu temple is better awning. The roof is a smaller and better -preserved
preserved. It consists of a porch, open hall example of the type used on the large temple; the
(rarigamal}(iapa) and sanctum door-frame. The same is true of the domical ceiling within. The porch
sanctum itself has completely disappeared. This and interior have highly ornate pillars embellished
temple is raised on an ornate socle (pifha) crowned with figures. This use of attached figures might seem
by running friezes of elephants and human figures to be a Kacchapaghata device , but a similar
(Fig. 11). Mouldings such as these are not seen configuration is found on one of the columns in
earlier in the Gwalior region. In the 9th century, Quwwat al-Islam mosque at Delhi (Fig. 12).
temples generally lacked any kind of plinth save for The idol chamber of the larger Sas Bahii temple
square slab-mouldings ( bhitta). Even at Sihoniya the has been virtually destroyed, though some modern
plinth is relatively low and is constituted so as to attempts have been made to reconstitute the outer
resemble a vedibandha (Fig. 2). The mouldings that walls in a rough-and-ready fashion. The cella of the
22 Afjchae/ D. Wi/Jjs

Fig. JO. Gwalior fort, larger Sas Bahii temple, detail of exterior wall, dated by inscription VS 1150/AD 1093- 94.

smaller Sas Bahii has, of course, completely the subcontinent (see Dirks, 1987; Inden, 1990, pp.
disappeared. The destruction of these parts is 228-62; Willis, 1993). From the 7th century, and
customarily written off as an example of religious more particularly in north India from the 10th, ruling
fanaticism. While it might seem relatively innocuous kings, as devotees, intermingled their personalities
to regard the Sas Bahii temples as expressions of with those of specific gods. In the first place, kings
Yai~Qava piety and their desecration a measure of shared their names with temple deities, a
religious intolerance, this kind of interpretation is a metaphorical likeness amplified by the eulogistic
secularized critique that separates religion and polity epigraphs which compared their attributes and
to the advantage of the former. At its worst, this achievements. A fuller sense of the relationship is
interpretation flies in the face of the inscriptional provided by the royal rituals carried out at
evidence and is an aggressive attempt to deprive Vijayanagara (Stein, 1983). Specifics of these rituals,
the monument of historical content. particularly the mahanavamf festival, were unique
While we cannot deal with these issues in detail to South India in the 15th and 16th centuries, but I
due to limitations of space, it is worth touching on do not think it is unreasonable to infer from this
the matter in view of its contemporary .relevance. later material that the Kacchapaghata king was the
The dedication of the larger Sas Bahii to principal devotee and that a centrepiece of ritual in
Padmanatha in honour of king Padmapala the larger Sas Bahii was the king's worship of
demonstrates that the Kacchapaghata kings equated Padmanatha. Extending this inference we might
themselves with the deity they had established in suggest that the king entrusted his regalia to the
the temple. In this the Kacchapaghatas were not god during worship and his annual lustration
alone , a fact documented by examples throughout ( abh1$eka). Such acts, or at least ones broadly similar
 Kacchapaghata Temples 23

Fig. 11. Gwalior fort , smaller Sas Bahii temple , plinth and parapet wall, datable by inscription VS 1150/AD 1093- 94.

to them, were the instruments through which the destructive acts must be put in the context of the
Kacchapaghatas constituted and maintained their Islamic practice of demolishing the monuments of
power. In this light , the motivation for the earlier Islamic regimes and the long history of
destruction of the Sas Bahfi sanctum is reasonably temple looting within the Indian tradition itself
clear. Defeat of the dynasty was signalled not only (Davis , 1993). ·
by military reversals in battle , but by the destruction The historical sketch given above illustrates the
of the institutions and ultimately the buildings centrality of the Sas Bahfi temples and shows why
through which the monarchs sought to construct they were singled out for destructive attention.
their legitimacy. The demolition of the sanctum and Other buildings were less important and are
the smashing of the deity quite literally decapitated consequently in a better state of preservation. The
the Kacchapaghata regime. The temple-hall at most instructive example is a ruined temple near
Gwalior thus represents a headless corpse , left the Suriij Kui:i<;I
on Gwalior fort (Fig. 13). This gives
behind as a poignant sign that the old dispensation some indication of the appearance of the missing
had been brought to an end. Perhaps nothing else parts of the Sas Bahfi temples. The lower mouldings
so clearly indicates that the arts of peace and war of the plinth (pifha) have been stripped away, but
were intimately connected , and that the historical the inverted cyma Uaifyakumbha), pointed arris
process through which this building has passed (kar.(11ka")and band of grotesques (grasapatt1) are
recapitulates the metaphysical truth, oft-heard in reasonably well preserved. Above these mouldings
India, that creation contains the seeds of destruction. rises a prominent socle ( vedibandha). The treatment
Certainly the desecration of temples in India is of these forms diverges significantly from Khajuraho
associated most commonly with Mal)mfid of Ghazni and indicates that distinct architectural practices
and his successors in the Delhi sultanate , but these were maintained in Gopak~etra under the
24 Michael D. Willis

and the depth of the recesses between them is again

at variance with Khajµraho standards. Particularly
telling is the fillet on the edge of the torus, a virtual
signature of the mature Kacchapaghata style. Above
the string-courses is the wall-section (Ja.ngha),
perhaps the most unusual feature of this building.
The degree of architectural clarity is particularly
evident when we juxtapose these plain walls against
the tightly packed tiers of sculpture in the Khajuraho
temples . The absence of sculpture allows the offsets
to be developed into pilasters, a theme that is
carried into the entablature ( vara1.u;Jikajwhere the
pendant-like leaves represent capitals . Also of note
in the entablature are the mouldings constituted as
a vedibandha. In 7th-century architecture, elements
of the podium were often reproduced in the upper
parts of the spire to represent an elevated altar or
uttaravedi, but the preservation of this theme at such
a late date is unusual. The superstructure of the
Gwalior temple has fallen, but it may be assumed
that it was some sort of multi-spired sekhari, perhaps
akin to the example surviving at Kadwaha .
Fragments which may have come from this
superstructure , or from others of the same period ,
can be found on Gwalior fort. The Kadwaha temple ,
it should be noted, alerts us to the fact that plain
walls were not a uniform practice in the 11th
century, but represented one of several options
which architects could pursue.
A number of shrines and temples in the
countryside around Gwalior are related to those just
discussed and further illustrate the variety of
architectural forms in Kacchapaghata territory. The
most interesting, at least among those I have been
able to locate , are found in the vicinity of Paroli:, a
village of considerable antiquity. It is perhaps not
out of place to note that ParoU (or Paroii) was
known anciently as Pac;lhavali, a name testified by
Fig. 12. Delhi , Quwwat al-Islam mosque , pillar , 11th
an unpublished hero -stone inscription (now in the
century.
Archaeological Museum , Gwalior). In the jungle
Kacchapaghatas. In the first case, the vedibandha- near ParoU is the site of Batesar , a name derived
mouldings do not carry image niches, but flaring from the old temple there dedicated to Batesar
appliques with spade-shaped tips. Nothing of this Mahadev. The path between Batesar and Paroli is
sort is seen in the Khajuraho temples. Secondly, the dotted with many ruined temples and a significant
foot-moulding is very tall and provided with a number of them belong to the Kacchapaghata
cornice-like cap. This recalls the lotus-cap employed period. A ruined hall is the most impressive of the
on the Kakanmath at Sihoniya. The kalasa (torus) group. It is cruciform in plan like the smaller Sas
is no longer prominent, but becomes one of a series Babu and has closely related pillars internally (Fig.
of string courses (maiicika") placed below the wall 14). The mouldings, however, are less elaborate
and image nich es . The variety of these mouldings while the parapet pickets (stambhika") and recessed
 Kacchapaghata Temples 25

Fig. 13. Gwalior fort, ruined temple near Suriij Kui:i<;I

, 11th century.

panels (phalaka) lack stencil-like patterns. The 8th and 9th centuries. The Kacchapaghata example
door is likewise less elaborate. These points suggest at Paroli retains the essential features of the type,
that the Paroli ruin is earlier than the Sas Bahu notably receding slabs and projecting pilasters; the
temples; a date in the mid-11 th century seems niches are also crowned by simple pediments
likely. ( udgama) in the traditional fashion. The pilasters,
Other buildings also document mid-l lth century however, are stripped of ornament and the
architecture , albeit at a more humble level. The pediments treated with greater flatness than in
structure illustrated here in Fig. 15 may b~ taken as earlier times. A second shrine a few kilometres
representative of the many shrines which have southeast of Paroli is of the same type (Fig. 16).
survived in the general vicinity of Paroli. This temple This building introduces guardian figures on the
illustrates the later history of a type now generally corners , an arrangement derived from the
known as maJJ(}ap1ka(Meister, 1976). While we shall iconographic programme of latina temples. A
use the term ma}J(iapika here , it must be noted that variation of thi~ is seen in a third shrine which has
the meaning of the word was not so restricted in unusual niches straddling each corner of the
epigraphic usage as it has recently become in art- structure. Directly beside this shrine is a large figure
historical parlance (see Sircar, 1966, s.v.). A mainstay of a Jaina tirthamkara , one of the best preserved
of shrine architecture in central India from the 7th 11th-century sculptures in the area (Fig. 17).
century, the maJJ(}apika-type was first characterized The subsequent development of ma}J(iapika
by walls composed of alternating pilasters and slabs. architecture is documented by a Vi$JJUtemple at
Perhaps derived from post -and-plank construction Tilori (Fig. 18). This building is preceded by a ruined
in wood, the ma,JJ(iapika enjoyed its heyday in the hall of which only the pillars and lintels remain ; one
26 Michael D. Willis

Fig. 15. Paroli , shrine of the mar:iQapikii, mid -llth

century.

Fig. 14. Paroli , ruined temple hall , interior , mid-llth

century.
(Fig. 19). That this temple is later than the
m8.IJ<;Japika-shrinesof the mid-llth century is not
pillar carries an inscription of a later date (Dvived1, likely to be argued. While the decorative surface is
YS 2004, no. 306). The temple has a large platform analogous to the Sas Bahii temples, the shrine has
(jagall), now partially disturbed. The exterior walls none of their considered architectural proportion or
of the shrine -proper are composed of a medley of decorative panache. There are , in fact , no
ornamental pilasters , chequered recesses and niches architectural features derived from the Sas Bahii
with pediments. The pilasters have the flat stencil- project; all the forms are drawn from a humble
like patterns of the larger Sas Bahii temple, while architectural repertoire that seems to have
the niches retain pediments of an older v·ariety. On maintained its basic vocabulary apart from the
the corners (kanJa) and vestibule wall (kapi/J) the ambitious monuments of the Kacchapaghata rulers .
niches have been cut to accommodate the required This indicates, in my view, something of the internal
space, a late and awkward rendition of the half- autonomy of ancient Indian society, where local
candrasiila motif that is found in temple architecture practices and traditions were allowed to continue
from the 7th century. The door of the Tilori shrine without interference from the centre , even a centre
continues the configuration seen earlier at Paroli so close as Gwalior.
 Kacchapaghata Temples 27

plinth. There may have also been a hall, but

excavation would be necessary to confirm this.
Superimposed sanctums are unusual but not
unknown in Indian architecture. The most famous
instance is the Vaikui:itha temple at Kanchi, but the
Sas Bahii. hall provides a closer parallel with its
door-like opening above the main entrance (Fig. 9).
A date in the 12th century for the ruin is suggested
by the general flatness of the offsets, the reduction

Fig. 16. Paroli, Siva temple, mid-llth century.

The dynastic history of the Kacchapaghatas after

the late-11 th century is unclear (Trivedi, 1980), but
enough is known to provide a tentative outline of
architectural developments. We have an inscription
of Ratnapala (c. 1105-30) mentioning the foundation
of a Siva temple, but this record has not been
properly edited (Sinha , 1980). When discovered, the
inscription was built into the curtain wall of the fort
near the Sas Bahii temples ; it was removed before
1980 to the Archaeological Survey of India stores
(near Man Mandir). The building to which this
inscription refers seems to have disappeared, but a
ruined structure of approximately the same period
is located at a central point on Gwalior fort. This
building, in its current state, consists of three
superimposed sanctums. It was probably crowned
by a multispired superstructure and surrounded by
an ambulatory resting on a substantial moulded Fig. 17. Paroli , standing tirtharilkara image, 11th century.
28 Michael D. Willis

Fig. 18. Tilori, Vi~i:iutemple, late 11th century.

of the door jambs to smooth panels and the absence inward to a cloistered court. In the centre of the
of niches with figural sculpture. Perhaps the most court is a circular shrine to the god Siva, also rebuil t
startling feature is the way in which the components in its upper parts. The lower portion shows a
of the elevation (mal)ifovara) have lost their moulded plinth and an open circular ambulator )
architectural integrity and been reduced to a series supported by paired columns. The individual pillar~
of moulded courses. have bases and double capitals, but lack carved flora
An analogous reductionism is seen in the ornament of any kind. The cells lining the courtyar d
circular temple at Mitaoli (Fig. 20). This building, are empty, but they probably contained yogin
probably also from the time of Ratnapala, has been images like the well-known temple at Bheragha t
heavily reconstructed, but its original appearance can (Sharma, 1978). Other buildings of this type an :
be discerned. The outer walls are provided with found at Khajuraho and Dudhahi (near Deogarh).
offsets, each carrying a small niche and pediment. The last phase of architecture under th e
The niches flanking the rebuilt entrance contain Kacchapaghatas is illustrated by the temples a
figures of Siva and Parvati; the others, where Naresar (ancient Nalesvara). This site is best know!'
preserved, have only diamond-shaped lozenges. for its 8th-century monuments, but additions were
Below the niches is a continuous base consisting of made in the 12th century as testified by severa
a podium ( vedibandha) and plinth (pi_tha). The inscriptions. The earliest record of the period is
mouldings are particularly sharp and compressed. dated VS 1202/A.D. 1145-46 and records a gift to a
Above the wall and its niches is a continuous temple by the son of Sri Haripfila. Not long afte1
entablature ( varapifika'). Enough is preserved above this are a group of mother-goddess images, now ii'
the entablature to show that the temple was once the Archaeological Museum, Gwalior. These are
ringed with small Jatina-shaped spires, something inscribed with the name of the donor ravala Sr
like the Adinatha temple at Ranakpur. The spires Vamadeva (Dvivedi, VS 2004, nos. 93 and 680-91)
at Mitaoli originally crowned small cells which faced in one case the date is given as VS 1245/A.D
 Kacchapaghata Temples 29

mouldings. Porches , when they do appear, are

rudimentary in the extreme . Extending trends
already apparent in the mid-llth century, the pillars
are devoid of relief scuplture and show only faint
vestiges of capitals and brackets. The Jatina-type
temples are equally austere. The spires still have
projecting faces but they lack carved ornament aside
from slender fluted discs (kar(lliJJ<;Jaka). This
development shows that by the late-12th century the
spire was on the verge of becoming a simple, smooth
turret. The large 12th-century temple at Naresar has
a superstructure in the same style (Fig. 22). The
lower portions , however , are more ambitious. The
podium ( vedibandha) and plinth (pftha) are provided
with recognizable mouldings and the wall with
images. In marked contrast with the past, the images
are not sheltered by pediments. The door has several
jambs and corkscrew pilasters, suggesting that it was
modelled on the Sas Bahii temple at Gwalior.
Indeed everything about these later shrines is
derivative; they represent the lingering vestiges of a
once great temple tradition which only a century
before had produced some of the most spectacular
buildings in the history of Indian architecture.
The sculpture of the late 12th century is marked
by an equally dramatic breakdown of traditional
practices and forms. Brittle modelling, awkward
Fig. 19. Tilori, Vi$1)U temple , entrance door, late 11th composition and misunderstood conventions
century. combine to create an art which seems to be tottering
on the verge of extinction . It was only in the 6th
1188-89. Another image-inscription gives the date century, after the collapse of the Gupta dominion ,
VS 1249 margas,r~a sudi 2 ravivara/ A.O. 8 that an analogous decline occurred. The goddesses
November 1192 (Ramasharma, 1969). This records from Naresar represent the case in point, but other
the establishment of an image, again by Vamadeva , examples are not wanting. Among the more intriguing
during the time of the Kacchapaghata ruler pieces, and one that takes us into the second half
Ajayapala. Vamadeva styles himself of the 13th century , is a free-standing pillar at Tilori.
Doc;lhak~etrapala indicating that he was a minor The round shaft of the pillar is interrupted by a
noble, most probably from Gwalior where a later rectangular section with image panels. The failing
inscription describes one of the gates there as ability of indigenous sculptors is neatly illustrated
Ohoc;lhapauri. Whatever Vamadeva's position, his by this uninspired work (Fig. 23). A long inscription
records allow us to assign most of the later buildings (Dvivedi, VS 2004, no. 619) of the ruler
at Naresar to the late-12th century. In all probability Kirtipaladeva is on the pillar but it bas not been
the images of the goddesses were installed in the studied despite the fact that it is far from clear how
shrines at the site. The buildings, of which there is the later Pratiharas (to whom the Tilor, pillar
a considerable number, carry pyramidal and apparently belongs) established themselves after the
curvilinear superstructures. The pyramidal or fall of the Kacthapaghatas. In these final years
phamsana types consist of simple chamfered slabs before the Kbalji conquest , only the Yajvapalas
of diminishing size (Fig. 21). The temple walls are of Narwar have received scholarly attention
without ornament and their bottom edge devoid of (Trivedi, 1959).
30 Michael D. Wi//is

Fig. 20. Mitiioli , Yogini temple , exterior , 12th century.

 Kacchapaghiita Temples 31

Fig. 22. Naresar , Siva temple , late 12th century. Fig. 23. Tilori , relief figures on pillar of Kirtipiila, 13th
century.

Conclusion Sihoniya is only symptomatic of a larger problem

of neglect. The time has past when we can consider
I have ended with the Tilori inscription to emphasize our work done having blithely cited Patil's inventory
the fact that this essay is not a comprehensive of sites (Patil, 1952) or Dvivedi 's list of inscriptions
history. l cannot claim to have covered all the (Dvivedi , VS 2004). Invaluable as these sources are ,
important monuments of the Kacchapaghata period they are based on reports written for the
and I have not tried to hide the deficiencies in the Archaeological Department of Gwalior State before
current state of our knowledge with regard to the World War II; many of the lesser-known sites and
documentation of buildings or the editing of monuments have not been visited by historians or
inscriptions. There are virtually hundreds of archaeologists for over half a century. A good case
buildings and dozens of inscriptions awaiting proper in point is Chait, a Jaina temple site southwest of
study and publication. The degree to which the Gwalior with several 12th-century temples. None
subject has been ignored is indicated by Sihoniya. have been published. What we urgently need is
This ancient capital has not merited an article, never thorough documentation if we wish to go beyond
mind a monograph, an astonishing situation if we the current restrictions within Indology and arrive
think of the amount of work done on cities of at a substantive understanding of this important
analo gous importance in medieval Europe. Alas period in the history of India.
32 Michael D. Willis

Acknowledgements lnden, R., 1978, Ritual, Authority and Cyclic Time in Hindu
Kingship, pp. 1-27 iri: Kingship and Authority in South
This article was first suggested to me by Krishna Asia, J.F. Richards, ed., Madison.
lnden, R. 1990. Imagining India, Oxford.
Deva whose interest in my work has encouraged Luard, C.E. 1908. Gwalior State Gazetteer, Bombay.
me to bring this and other projects to conclusion. I Meister, M. 1976. Construction and Conception: Maooapika
would also like to thank H. V. Trivedi, with whom Shrines of Central India, East and West, 26: 409-4 18.
I have enjoyed lively correspondence regarding the Patil, D.R. 1952. The Descriptive and Classified List
archaeology and inscriptions of the region. H.N. of Archaeological Monuments in Madhya Bharat,
Dvivedi and R.N. Misra not only shared their wide Gwalior.
Ramasharma. 1969. Two Inscriptions of Ajayapaladeva.
historical knowledge but opened their homes to me Epigraphia Indica, 38: 132-134.
during my long stays in Gwalior. Rowland, B. 1956. The Art and Architecture of India, 2nd
edition, Baltimore.
Sharma, R.K. 1978. The Temple of Chaunsatha-Yogini at
BIBLIOGRAPHIC REFERENCES Bheriighiit, Delhi.
Sinha, V.K. 1980. Kacchapaghiita riijavarilsa kii navina
Archaeological Department, Gwalior State, 1937-43, Annual abhilekh, Gvaliyar darsan, Gwalior, pp. 217-219.
Report, Sarilvat 1980, Year 1925-26 to Sarilvat 1997, Sircar, D.C. 1966. Indian Epigraphical Glossary, Delhi.
Year 1940-41, Gwalior. Srinivasan , P.R. 1970. Five Gangola Tank-Bed Inscriptions.
Archaeological Department, Gwalior State, 1949, Epigraphia Indica, 38: 305-312.
Quinquennial Administration Report for Samvats 1998- Stein, Burton. 1983. Mahanavami: Medieval and Modern
2002 (Years 1942-46), Gwalior. Kingly Ritual in South India, pp. 67-90 in: Essays on
Babb, L.A. 1993. Monks and Miracles: Religious Symbols Gupta Culture, Bardwell L. Smith, ed., Delhi.
and Images of Origin among Osviil Jains, Journal of Trivedi, H.V. 1953. The Bibliography of Madhya-Bharata
Asian Studies , 52: 3-21. Archaeology, Gwalior .
Brown, Percy. 1944(?]. Indian Architecture, 2 vols., Bombay Trivedi, H.V. 1959. The Yajvapalas of Narwar. Journal of
(with several later editions). the Madhya Pradesh Itithasa Parishad, 1: 22-32.
Davis, Richard H. 1993. Indian Art Objects as Loot. Joumal Trivedi, H.V. 1979-91. Inscriptions of the Paramaras,
of Asian Studies, 52: 22-48. Chandellas and Kachchapapaghatas and two Minor
Deva, Krishna. 1959. Temples of Khajuraho, Ancient India, Dynasties, Corpus Inscriptionum Indicarum, volume 7
15, pp. 43-65. (in 3 parts), New Delhi.
Deva, Krishna. 1963-64. Kachcchapaghata Temples. The Trivedi, H.V. 1980. Kacchapaghiita naresoril ke anuvarilsaj,
Researcher, 3-4: 5-9. Gvaliyar darsan, Gwalior, pp. 214-216.
Deva, Krishna. 1969. Temples of North India, New Delhi. Trivedi, R.D. 1990. Temples of the Pratihiira Period in
Deva , Krishna. 1987. Khajuraho, New Delhi. Central India, New Delhi.
Dirks, Nicholas B. 1987. The. Hollow Crown: Ethnohistory Watters, Thomas. 1904. On Yuan Chwang's Travels in India,
of an Indian Kingdom, Cambridge; 2nd ed., 1993, Ann 2 vols., London.
Arbor. Willis, M. 1988. An Introduction to the Historical
Dvivedi, Harihar Nivas. VS 2004. Gvaliyar rajya ke Geography of Gopak$etra, Dasan:ia and Jejakadesa.
abhilekh, Baniiras. Bulletin of the School of Oriental and African $tudies,
Dvivedi, Harihar Nivas. 1980. Gopak!;letra ke kacchapaghata, 51: 271-278.
Gvaliyar darsan, Gwalior, pp . 186-216. Willis, M. 1993. Religious and Royal Patronage in North
Handiqui, K.K. 1968. Yasastilaka and Indian Culture, or India , pp. 48-65, in: Gods, Guardians and Lovers,
Somadeva's Yasastilaka and Aspects of Jainism and Vishakha Desai and D. Mason, eds., New York.
Indian Thought and Culture in the Tenth Century, 2nd Willis M. 1996. Inscriptions of Gopaksetra: Materials for
edition, Sholapur. the History of Central India, London.