M a p p in g M a ja p a h it :

W a r d e n a a r ' s A r c h a e o l o g ic a l
Survey a t T r o w u l a n in 1815

Amrit Gomperts, Arnoud Haag,

and Peter Carey1

Old maps may exist for regions that identify now "lost" archaeological sites, and
georeferencing will help archaeologists to relocate those sites.
— Sarah H. Parcak, Satellite Remote Sensing for Archaeology, 2009, p. 50.

On August 22, 1815, the lieutenant-governor of Java, Thomas Stamford Raffles (in
office, 1811-16), ordered the army surveyor Captain Johannes Willem Bartholomeus
Wardenaar (1785-1869) to make a plan of the site of the ancient capital of Majapahit at
Trowulan (Illustration l).2Wardenaar produced a map, a map legend containing brief

1The authors would like to thank Mr. T. Richard Blurton (Curator for South and Southeast Asia
Collections, British Museum), Professor John Bastin (formerly of the School of Oriental and African
Studies, London), Dr. Roy Jordaan, Dr. Annabel Teh Gallop (British Library), Drs. Aris Soviyani
(Departemen Kebudayaan dan Pariwisata, Trowulan), Professor Ben Arps (Leiden University), Professor
John Miksic (University of Singapore), Ms. Pauline Lunsingh Scheurleer (formerly Rijksmuseum,
Amsterdam), and Ir. Heru Joko Susilo for their kind assistance with our research and the writing of this
article. We are also indebted to the International Institute of Asian Studies in Leiden, in particular to Dr.
Philippe Peycam, Dr. Manon Osseweijer, Professor Max Sparreboom, and Professor Wim Stokhof, for its
grant for the procurement of a high-resolution satellite image. This article is based on the present authors'
article: A. Gomperts, A. Haag, and P. Carey, “De Veertiende-Eeuwse Javaanse Hofstad Majapahit alsnog
op de Kaart Gezet,” Caert-Thresoor 27, 3 (2008): 71-78. The observations on site at Trowulan were carried
out in April and December 2008.
2 F. de Haan, "Personalia der Periode van het Engelsch Bestuur over Java 1811-1816," Bijdrngen tot de Taal-,
Land- en Volkenkunde 92, 4 (1935): 662.

Indonesia 93 (April 2012)

178 Am rit Gomperts, Arnoud Haag, Peter Carey

descriptions of the monuments, and several drawings. He sent the originals to Raffles
in late 1815 but did not keep a copy of the map.3 This map was subsequently
considered lost. Wardenaar's own copy of the legend and three sheets with drawings
were handed over to the Bataviaasch Genootschap van Kunsten en Wetenschappen
(Batavian Society of Arts and Sciences) in 1879.4The text of both Wardenaar's legend
and his jotted notes on his drawings have been published by the mining engineer and
pioneer of modern volcanology, Rogier Verbeek (1845-1926).5 Wardenaar's Plan of
Majapahit was subsequently traced by the present authors to the Drake Collection of
the British Museum.6This collection also includes copies of a few plates sent by
Wardenaar to the British lieutenant-governor.7 In this article, we will discuss the
importance of the 1815 map for present-day archaeologists of the Trowulan site.

Wardenaar's Archaeological Survey of Trowulan in 1815

On the eve of the colonial annexation of the area, Lieutenant H. G. Jourdan
prepared a report of the Wirasaba and Japan districts, dated April 28,1813, in which he
described the site of the medieval court-city:
Majapait, the ancient capital of the eastern empire is situated in Wirosobo, at the
distance of about eight or ten miles. The site is now overrun with a thick bamboo
jungle and few traces of this once extensive city are now to be discovered: idols,
which prove that the art of sculpture had attained some degree of perfection have
been dug up at different times near the spot, and there are still the remains of an
edifice evidently of Hindoo construction. The natives retain a kind of superstitious
reverence for these relics, which they are unable to explain but do not attempt to
conceal. They regard the idols as representations of evil genie whose favor it is
necessary to conciliate.8
According to the dates on his drawings, Wardenaar was at Trowulan between October
5 and 7, 1815. However, it must have taken the engineer captain several weeks to
survey the extensive Trowulan site, which at the time was covered by thick bamboo

3 W. R. van Hoevell, Reis over Java, Madoera en Bali in het Midden van 1847, vol. 1 (Amsterdam: Van Kampen,
1849-54), p. 184.
4 Notulen van het Bataviaasch Genootschap 17 (1879): 111. The documents were kept in the archives of the
Oudheidkundige Dienst (Archaeological Service) at Batavia in 1941. See W. F. Stutterheim, De Kraton van
Majapahit, Verhandelingen van het Koninklijk Instituut voor de Taal-, Land- en Volkenkunde van
Nederlandsch-Indie 7 ('s-Gravenhage: Nijhoff, 1948), p. 1, n. 3. But we have not been able to trace them in
Jakarta.
5 R. D. M. Verbeek, "De Oudheden van Madjapahit in 1815 en 1887," Tijdschrift voor Indische Taal-, Land- en
Volkenkunde 33 (1890): 1-15.
0 The present authors were able to consult Wardenaar's Plan of Majapahit at the British Museum on March
12, 2008. See http:/ /www.britishmuseum.org/collechonimages/ AN00477/ AN00477148_001_l.jpg
(Registration Number: 1939,0311,0.5.36). See also Gomperts et al., "De Veertiende-Eeuwse Javaanse
Hofstad Majapahit," pp. 74—75, 77.
7J. Bastin and B. Brommer, Nineteenth Century Prints and Illustrated Books of Indonesia (Utrecht: Spectrum,
1979), p. 340. Mrs. J. H. Drake, Raffles's great-great niece, donated Wardenaar's map to the British
Museum in 1939, along with a number of other manuscripts relating to Javanese antiquities that Raffles
had acquired during his lieutenant-governorship of Java (1811-16).
8 Lt. H. G. Jourdan, manuscript, "Report on Japan and Wirosobo," dated April 28,1813 (London: British
Library, India Office Library & Records, Mackenzie Private collection 21, pt. 10), p. 355.
 Mapping Majapahit 179

and teak forests. As a military surveyor, he would have had use of contemporary
instruments such as a surveyor's astrolabe, a plane table with an alidade, a compass, a
sextant, and chains.9The Plan of Majapahit at the British Museum is drawn on a scale of
1:12,000, and measures 42 by 35 centimeters. The spelling of names on the map
suggests that this plan is an English copy, probably dating from late 1815 or early
1816.10
The geometrical faithfulness and planimetric accuracy of Wardenaar's original
survey was evident to the present authors when we imported a digital scan of the Plan
of Majapahit onto Google Earth and projected it as a half transparent overlay on top of a
high-resolution satellite image.11 Using GPS mapping software, we were able to
georeference the Plan of Majapahit with coordinates of ground-control points taken with
GPS receivers, thus verifying the positions of the remains on site (Illustration 2). These
ground-control points and site identifications are marked in the present article with
angle brackets <...>. The corresponding coordinates are listed in the Appendix. The
absolute positioning accuracy of the plan is 54 meters with respect to ground truth.12
The georeferencing also shows that the scale bar on the plan indicates Rijnlandse Roeden
(i.e. Rhineland Rods) even though the plan itself states "English] Roods."13

9 R. F. de Seijff, "Overzigt der Geographische en Topographische VerrigtingenGedurende onze

Heerschappij in den Indischen Archipel," Natuurkundig Tijdschrift voor Nederlandsch Indie 9 (1856): 353.
Compare A. H. A. Hogg, Surveying for Archaeologists and Other Fieldworkers (London: Croom Helm, 1980).
Wardenaar studied at the Marineschool—the Naval College (1782-1812)—at Semarang from 1798 to 1805,
and he graduated as an ensign. See De Haan, "Personalia der Periode van het Engelsch Bestuur over Java
1811-1816," p. 661; and F. C. Oosten, "'Hear Instruction and Be Wise'; The History of the Naval College on
Java in the Eighteenth and Early Nineteenth Century," Mariner's Mirror 55 (1969): 247-62. Teachers and
pupils at the college specialized in training as marine hydrographers for VOC (Vereenigde Oost-Indische
Compagnie, Dutch East Indies Company) service, producing all the Dutch hydrographic charts of Asian
waters in the post-1782 period. See further C. J. Zandvliet, "VOC Maps and Drawings,"
http: //www.tanap.net/content/voc/maps/maps_making.htm, accessed February 15, 2012.
10 On the British Museum Plan o f Majapahit, "Trowulan" is given in the English spelling, but, both in the
Dutch and English texts of Wardenaar's legend, the "u" is spelled according to contemporary Dutch
language norms: "Trowoelan." See Verbeek, "De Oudheden van Madjapahit in 1815 en 1887," pp. 3-6. For
further cartographical arguments, see Gomperts et al., "De Veertiende-Eeuwse Javaanse Hofstad
Majapahit," p. 74.
11 The Quickbird satellite image (resolution: 0.6 meter per pixel) available on Google Earth was taken on
September 24, 2003 (scene number: 1010010002503301). The image was taken just after sugar cane was
harvested. Cane is grown in major parts of Trowulan area, so that during the harvest, which occurs every
eighteen months, the bare soils are visible.
12 For the georeferencing of Plan of Majapahit depicted in Illustration 2, we used GPS mapping software
(OziExplorer version 3.96.1c) and the GPS co-ordinates of nine ground control points <A, C, F, G, H, 1.1, L,
M.3 and N.l> with a first-order polynomial (i.e. affine) transformation. Our georeferencing of
Wardenaar's plan shows no skew. The orientation of the north arrow on the georeferenced plan (359.5°) is
consistent with the magnetic field declination for the region in 1815 (i.e. 0.5° west, see http:/ /geomag.org/
info/declination.html). However, the easting axis appears 3.8 percent compressed in comparison to the
northing axis. For the estimation of the overall planimetric accuracy of the 1815 plan, we used 18 ground
control points <A, B, C, D, E, F, G, H, 1.1, K.l, L, M.l, M.2, M.3, M.4, N.l, N.2 and N.3>, their
corresponding pixel positions in x and y coordinates on the digital scan available at the website of the
British Museum, which has a resolution of 300 dots per inch (corresponding to 2.6 meters per pixel), and a
least squares algorithm for first-order polynomial transformations. On Plan of Majapahit, the planimetric
residuals in the northings (standard deviation: 22 meters; maximum residual: 50 meters) of the features are
larger than in the eastings (standard deviation: 14 meters; maximum residual: 26 meters). The maximum
circular residual amounts to 54 meters <E>.
13 Before the introduction of the metric system in 1816, the Dutch used measures referring to the Rijnland
system in the Netherlands East Indies. By royal decree of February 18,1808, one Rhineland Rod (Dutch:
180 Am rit Gomperts, Arnoud Haag, Peter Carey

Wardenaar's map legend contains fifteen items labeled A to P, with those listed A
to H rendered in both Dutch and English. The remainder appear only in Dutch. For
convenience, we present only the English version of items A to H here. We have
translated the Dutch texts of legend items I to P and the notes on the drawings as
published by Verbeek.14We have not adopted Wardenaar's alphabetical order A to P,
but instead follow a spatial arrangement in separate sections of the former royal
capital. Nearly all the remains shown on Wardenaar's drawings are still extant and
have been depicted in various publications. In the present article, therefore, we include
only three plates: the site of M6nak Jinggci <K.1>, copies of Wardenaar's plates of the
gate of Bajangratu <H>, and the vanished Candi Muteran <A> (Illustrations 3-5).

The Vanished Fourteenth-Century Royal Palace of Majapahit

On the basis of what he knew about traditional Central Javanese courts, Wardenaar
infers that the position of the vanished royal palace (kedaton) must have been to the
south of the hamlet of Kedaton and the adjacent Siti Inggil (Sitinggil) terrace:

N. "An elevated place named Siti Inggil—"high ground" usually in front of the place of
the king's palace. It is high ... [illegible number] feet and [its sides] lined with baked
bricks. On one of the corners there still stands a part of a pillar of the same bricks,
probably as footings for wooden railings with which one might surmise it was [once]
surrounded. The small red dot therein is a dry well [measuring] 1 foot 10 inches [58 cm]
wide, 2 feet 9 inches [86 cm] long and I 8 V2 feet [5.8 m] deep, called Sumur Windu by the
[local] inhabitants."
(sub " ... and one might therefore surmise that, in accordance with the lay-out generally
P.) adopted by the indigenous people, the king's palace would have been situated to the
south of the aforementioned Siti Inggil, around the hamlet of Kedaton, whose name
also denotes a king's residence."

In fact, Wardenaar's plan shows two adjoining hamlets of Kedaton extending over
a larger area than the present-day settlement first shown on the 1879 topographical
map. The hamlets exactly coincide with the area of the vanished kedaton, or "royal
palace," in W. F. Stutterheim's archaeological reconstruction.15Moreover, the Dutch
architect and archaeologist Henri Maclaine Pont (1884—1971) was able to observe
important remains of walls in 1925-26, which appear as the outer and inner palace

Rijnlandse Roede) converts into 3.767358 meters. One hour walking (Dutch: een uur ganns, French: mille
Hollnndnis) corresponds to 1,200 Rhineland Rods. One Rhineland Rod equals 12 feet (Dutch: voeteri). There
are 12 inches (Dutch: duimen) in a Rhineland Foot. See G. J. Knaap, J. R. van Diessen, W. Leijnse, and M. P.
B. Ziellemans, Grote Atlas van de Verenigde Oost-Indische Compagnie/Comprehensive Atlas of the Dutch United
East India Company, Volume II: Java en Madoera/Java and Madura (Voorburg: Asia Maior, 2007), pp. 35, 352.
Moreover, on the georeferenced plan depicted in Illustration 2, the northing axis is scaled 1:12,100.
Therefore, we conclude that Wardenaar intended to draw his Plan of Majapahit in a scale of 1,000 feet per
inch corresponding to map scale 1:12,000.
14Verbeek, "De Oudheden van Madjapahit in 1815 en 1887," pp. 3-7. The legend and the notes on the
drawings refer to the versions that Wardenaar kept for himself. We have not been able to trace a legend in
Raffles's collection.
15 Gomperts et al., "De Veertiende-Eeuwse Javaanse Hofstad Majapahit," p. 73, figure 5; and A. Gomperts,
A. Haag, and P. Carey, "Stutterheim's Enigma: The Mystery of his Mapping of the Majapahit Kraton at
Trowulan in 1941," Bijdragen tot de Taal-, Land- en Volkenkunde 164, 4 (2008): 411-30.
 Mapping Majapahit 181

walls in Stutterheim's archaeological reconstruction but which have all now

vanished.16
The four sides of the rectangular Siti Inggil terrace were lined with a now vanished
brick course. On the georeferenced map, the platform measures 55-60 meters east-west
by 65-70 meters north-south. This corresponds to the actual situation on site <N.2, N.3,
N.4>. Parts of the southeastern and the northwestern sides of the terrace have also now
vanished. The terrace is about two meters higher than the surrounding area. On
Wardenaar's plan, a footpath leads over the area of the Siti Inggil terrace, from the
middle of the north side to the middle of the south side, and passes a site now known
as Sumur Upas.17
Sumur Windu <N.1> literally means "Walled Well" and is a reference to the brick
walls that line the interior of the well. Wardenaar's description "small red dot,"
referring to Sumur Windu <N.1>, is represented by a small black rectangle <N.1> on
the plan at the British Museum. Situated near the northeastern corner of the Siti Inggil
terrace, it currently measures 1.50 meters north-south by 1.35 meters east-west, with a
depth of some 7 meters to ground water level. When the Dutch notary Nicolaas van
Meeteren Brouwer (1798-1831) visited Trowulan on June 21, 1825 just before the
outbreak of the Java War, he noted that local tradition related that the last Majapahit
king had vanished through this well.18Furthermore, 4 meters to the east of Sumur
Windu, there is a brick platform now called Candi Kedaton, which is also aligned on a
north-south axis and measures 12.60 meters by 8.50 meters with a height of 1.58
meters, facing west.19According to J. Knebel, a member of the Oudheidkundige Dienst
(Netherlands East Indies Archaeological Service) who interviewed Mangoen Amidjaja,
the Trowulan villager who was the guardian of these remains in 1907, a local tradition
describes how the Majapahit kings used this platform as their royal seat while granting
audiences to their senior officials.20 From Kromodjojo Adinegoro's 1899 report, we can
infer that Candi Kgdaton was still covered by soil in 1815.21 However, it seems unlikely
that a well dating from this period would have been situated just four meters in front
of such an important edifice as the Candi Kedaton.

16 H. Maclaine Pont, "De Historische Rol van Majapahit: Een Hypothese," Djihva 6, 4-6 (1926): 294-317; see
map Majapahitsche Restantenkaart (Map of Majapahit Remains).
17 Sumur Upas is the name of a dry well that is situated some 24 meters to the south-southwest of Sumur
Windu <N.1>. Wardenaar's mapping is sufficiently detailed to enable us to conclude that he did not
confuse Sumur Upas with Sumur Windu.
18 N. van Meeteren Brouwer, "Dagverhaal van eene Reis door den Oosthoek van Java, in het Jaar 1825,"
Mnemosyne, Mengelingen voor Wetenschappen en Fraaije Letteren 8 (1828): 304-5.
19 Mundardjito, ed., Rencana Induk Arkeologi Bekas Kota Kerajaan Majapahit, Trowulan (Jakarta: Departemen
Pendidikan dan Kebudayaan, 1986), p. 131.
20 J. Knebel, Rapporten van de Commissie in Nederlandsch-Indie voor Oudheidkundig Onderzoek op Java en
Madoera, 1907, Bataviaasch Genootschap van Kunsten en Wetenschappen (Batavia/The Hague:
Albrecht/Nijhoff, 1909), p. 66. See Gomperts et al., "Stutterheim's Enigma," p. 419, figure 5.
21 R. A. A. Kromodjojo Adinegoro, "Rapport van den Regent van Modjokerto en Djombang aangaande
diens Voorloopig Onderzoek van Oudheden in het Gehucht Kedaton (Desa Trawoelan), Residentie
Soerabaja," Notulen van het Bataviaasch Genootschap 37 (December 1899): CIV-CVI, app. 16.
182 Am rit Gomperts, Arnoud Haag, Peter Carey

The Alun-alun Square

The Dutch word passeerbaan is a corruption of the Javanese word paseban denoting
"a place where serving duties are performed while seated."2223The latter could refer to a
building, a yard, or a square. However, in the period 1600-1850, the Dutch used the
word passeerbaan to refer specifically to the large square, or alun-alun, in front of courts
and princely residences of Java. This is apparent from Stutterheim's convincing
explanation of the etymology of the Dutch-Indies passeerbaan22 and from Javanese and
Madurese town plans of the VOC period.24256Wardenaar's description clearly refers to a
space, not a building:

P. "A barren and flat space of ground covered only with grass and a few small shrubs,
from which one considers that the Passeerbaan was once situated here."

Given Wardenaar's description and the strict meaning of the Dutch colonial word
passeerbaan, it is beyond doubt that the oral tradition refers to the location of the alun-
alun square <P.1> of the now vanished royal capital. However, within the immediate
vicinity of Wardenaar's point <P.1>, Indonesian archaeologists have excavated the
remains of medieval dwellings <P.2, P.3, P.4>. From our observations of the soil
profiles on site, supported by detailed cartographical analysis, we conclude that the
soils in an area immediately to the east of these remains have been excavated by
colonial and present-day asset strippers to depths of 2-3 meters. There is more
evidence pointing to the location of the alun-alun <P.1> in the Balinese
historiographical chronicle Kidung Pamancangah.2-

The Tank of Segaran

Raffles describes the tank of Segaran as follows:
Here the walls of the tank, upwards of a thousand feet in length and not less than
six hundred in breadth, are quite perfect. They are of burnt brick and about
twelve feet high. The whole area of the tank, when I visited it [in May 1815], was
one sheet of beautiful rice cultivation, and almost surrounded by a noble forest of
teak.2b
Wardenaar confirms that the tank was used for rice cultivation in 1815:

22 Stutterheim, De Kraton van Majapahit, p. 55.

23 Ibid., pp. 55, 57.
24 Compare Knaap et al., Grote Atlas van de Verenigde Oost-Indische Compagnie, pp. 362-64, 377-78, 391-92,
401-2, 407, 410-11, 429.
25 See further A. Gomperts, "More on the Archaeological Mapping of the Fourteenth-century Royal City of
Majapahit Based on a Balinese Description in the Kidung Pamancangah," in From Beyond the Eastern
Horizon: Essays in Honour of Professor Lokesh Chandra, ed. Manju Shree (New Delhi: Aditya Prakashan, 2011),
pp. 63-73. See also A. Gomperts, A. Haag, and P. Carey, "Rediscovering the Royal Capital of Majapahit,"
Newsletter of the International Institute for Asian Studies (HAS) 53 (Spring 2010): 13, figure 3 #3 (available
online at http: //www.iias.nl/ files/IIAS_NL53_1213.pdf).
26 T. S. Raffles, The History of Java, vol. 2,1978 reprint (Oxford: Oxford University Press, 1817), p. 54.
 Mapping Majapahit 183

M. "A dug-out tank lined with bricks called Segaran or the "Little Sea," people say that it
was once 1814 feet [5.8 meters] deep with a floor of similar bricks, but now it is not more
______ than 9 feet [2.8 meters] deep and the Trowulan villagers plant their rice here."__________

On Wardenaar's plan, the tank of Segaran measures 365 meters long and 170 meters
wide. This corresponds to the measurements given by Resident Hendrik Jacob Domis
(1782-1842, in office, 1831-34) in his 1834 published report, in which his measurements
are 1,200 feet (377 meters) by 545 feet (171 meters).27The tank was restored in the 1980s.
Although Wardenaar's plan does not show a stream or canal feeding the rice paddies
in the tank, Van Meeteren Brouwer confirmed the existence of such a stream in July
1825.28 Indeed, since no bridges or crossing planks are marked on the engineer
captain's 1815 plan, it seems clear that he excluded hydrological features of the
Trowulan area when he did his survey.
Wardenaar describes a speelhuisje, a word that translates as "a small pleasure house
in a garden:"2930

O. "An elevation—here are several stone footings for wooden pillars belonging to a house,
one says that here the king's small pleasure/garden house (speelhuisje) once stood, there
are also two stone pillars here, each approximately 8 feet [2.5 meters] high and 7 inches
[18 cm] in diameter, [which] local inhabitants say would have been used for attaching
the elephants."

On site, three of these "elephant" pillars are still standing <0.2, 0.3, 0.5>, while
the inhabitants indicated to us the spot where a fourth pillar once stood <0.4>. The
four pillars mark the corners of an oblong-shaped area measuring some 75-77 meters
north-south and 22-33 meters east-west. However, the dimensions of the pillars differ
from Wardenaar's measurements: their height above ground is at present about 1.80
meters and the diameter of their irregular pentagonal to hexagonal sections range
between 25 and 37 centimeters. Furthermore, in the center of the area marked by the
four "elephant" pillars, at the spot of the speelhuisje, a small platform may have once
stood on a slightly elevated spot, approximately one meter higher than the
surrounding area < 0 .1 > . According to the reverend J. F. G. Brumund (1 8 1 4 -6 3 ), 30 two
inscribed stones showing the eight-pointed symbol of the Majapahit aureole, the
heraldic emblem of the Majapahit royal family, were taken from this area near the
Sggaran tank to the house of the Dutch Assistant-Resident in Mojokgrto. On site, we
saw a few other small shaped rocks and footings within a radius of 30 meters from the
elevated spot <0.1>, stones, which may have once belonged to the platform where the

27 H. J. Domis, "Bijzonderheden betreffende Sourabaya en Madura," De Oosterling, Tijdschrift bij Uitsluiting

Toegewijd nan de Verbreiding der Kennis van Oost-Indie 2 (1834): 90.
28 Van Meeteren Brouwer, "Dagverhaal van eene Reis door den Oosthoek van Java, in het Jaar 1825," p.
304.
29 W. Sewel and E. Buys, Voiko men Woordenboek der Nederduitsche en Engelsche Talen/A Compleat Dictionary of
Dutch and English, vol. 2 (Amsterdam: Cornells de Veer, 1766), p. 737.
30 J. F. G. Brumund, Indiana: Verznmeling van Stukken van Onderscheiden Aard over Landen, Volken, Oudheden
en Geschiedenis van den Indischen Archipel, vol. 2 (Amsterdam: VanKampen, 1854), p. 112.
184 Am rit Gomperts, Arnoud Haag, Peter Carey

Majapahit kings held seated audience with their prominent officials, thus enhancing
the royal character of the spot.
Elephants are Indian symbols of kingship. The Sanskrit text Arthasastra 2.31.3 (c.
third century CE) and the medieval Sanskrit architectural text Mayamata 29.171-175
reveal that the Javanese practice of keeping elephants with their feet attached to such
stone pillars originates from India.31Textual comparison between stanza 11.2 in the
Old Javanese Ramayaya Kakawin and the parallel lines (10.9) in its Sanskrit prototype
Bhaftikain/a further reveals that such poles for attaching elephants (Sanskrit alana) must
have already existed in Java at the time of the composition of the Old Javanese text,
probably in the early tenth century CE.32Furthermore, in the book of the second Dutch
navigation to the East Indies in 1598-1600, there is a drawing that shows the king of
Tuban seated on a small platform and surrounded by his nobles and tethered
elephants during a royal audience with Dutch merchants on January 24, 1599.33The
travel account both describes and depicts the feet of the royal elephants as being
tethered with chains to the base of a pillar. Therefore, oral tradition may have correctly
preserved the original purpose of the pillars near the speelhuisje. Furthermore, in his
Old Javanese dictionary, P. J. Zoetmulder notes that the word palimanan refers to a
"place where the king is seated in audience; probably not the seat itself; a pavilion-like
construction or dais? The connection with liman [elephant] is not clear."34 Limitations
of space do not allow us to present further arguments regarding the identification of
the pillars and speelhuisje as the former Majapahit palimanan.35 But it is quite possible to
imagine the role played by the platform in relation to the tank. Majapahit kings would
have displayed their kingship as symbolized by the four elephants while watching
festivities performed on the great expanse of the Segaran tank.36

Gajah Mada's Residence

Wardenaar's siting of the residence of the famous Majapahit prime minister (patih)
Gajah Mada (in office, AD 1331-64), is unique inTrowulan history:

31 B. Dagens, Mayamata: An Indian Treatise on Housing, Architecture, and Iconography (New Delhi: Sitaram
Bhartia Institute of Scientific Research, 1985), pp. 277-78.
32 S. Santoso, Ramayana Kakaivin, vol. 2 (New Delhi: International Academy of Indian Culture, 1980), p. 289.
Compare O. Fallon, Bhatti's Poem: The Death ofRavana, The Clay Sanskrit Library (New York, NY: New
York University Press/J JC Foundation, 2009), p. 226-27.
33J. Keuning, De Tweede Schipvaart der Nederlanders naar Oost-Indie onder Jacob Cornelisz. van Neck en Wybrant
Wamhjck 1598-1600, De Linschoten-Vereeniging, no. 46, vol. 3 (The Plague: Nijhoff, 1942), p. 177, figure 20.
34 P. J. Zoetmulder, with the collaboration of S. O. Robson, Old Javanese-English Dictionary, Koninklijk
Instituut voor Taal-, Land- en Volkenkunde, vol. 1 (The Plague: Nijhoff, 1982), p. 1029.
35 See further Gomperts, "More on the Archaeological Mapping of the Fourteenth-century Royal City of
Majapahit," pp. 69, 72.
36 "We discover neither in Java, nor in any other country of the Archipelago, any of those enormous tanks
of the southern part of Hindustan, on which the agriculture of whole provinces entirely depends." J.
Crawfurd, History of the Indian Archipelago, vol. 1 (Edinburgh: Archibald Constable, 1820), p. 352. In the
tank of Segaran, the stored amount of water would have been too small for irrigation purposes, at most
supplying an area of some 20 hectares of sawah. The abundance of ancient wells and the present-day use of
shallow groundwater in the Trowulan area for domestic water supply, easily replenished by wet season
rainfall and possibly inflow from the mountainous areas in the south, indicate that, in Majapahit times,
potable water was most likely supplied from wells. However, the tank of Segaran may have been used as
a supplementary source of water in case of drought or blocked canals or rivers.
 Mapping Majapahit 185

G. "A gate called by the inhabitants Gapura Jatipasar, almost ruined, as to be seen in the
drawing."
Note in ink on the accompanying drawing: "Gapura or gate of Jatipasar in the forest of
Majapahit. Situated to the northeast of [the] Segaran. The orientation of the entrance is
east-west, one says that this was the gate to the dalem of the pepatih Gajah Mada.
Surveyed on the 7thof October 1815 on the orders of Lieut.-Governor T.S. Raffles by
Wardenaar."
Additional note in pencil: "Gate or Gapura at Jatipasar in the forest of Bajangratu,
situated about half an hour walking to the northeast of Majapahit. The orientation of the
entrance is east-west. Surveyed on the 7thof October 1815."

The gate near the village of Jatipasar became known as Candi Wringinlawang by the
end of the nineteenth century. It is a captfi bentar, or split gate. Since its restoration in
the 1990s, the two parts of the gate are now of equal height, 15.50 meters high,37but in
the nineteenth century, the remains of the northern and southern parts were,
respectively, 48 feet (15.1 meters) and 21 feet (6.6 meters) high.38 Moreover, H.
Maclaine Pont observed that the split gate formed the southwestern entrance to a
complex. Its surrounding walls, measuring approximately 200 meters in an east-west
axial direction and 150 meters north-south, have all now vanished.39Furthermore, in
Mundardjito,40 a detailed plan reveals thirteen medieval wells 200 meters to the
southwest of the split gate within an area covering less than half of a hectare. So many
wells in such a small area indicate a dense population living on the western side of the
split gate in Majapahit times, but only four of these wells survive today. While Krom41
argues that the cat.nfi bentar is an architectural feature unknown in the palaces or
residences of pre-colonial Java, Wardenaar's oral tradition testimony directly
contradicts this. Since the engineer captain was almost certainly able to converse with
the local Trowulan villagers in his mother tongue, Javanese, his testimony strikes us as
being the more credible.42From topographic interpretation of Prapanca's description of
Majapahit in Nagarakrtagama 12.4 (1365 CE), we infer that Gajah Mada's residence was
situated in close proximity to the Candi Wringinlawang.43

37 Mundardjito, ed., Reneana Induk Arkeologi Bekas Kota Kerajaan Majapahit, p. 166.
38 Van Hoevell, Reis over Java, Madura, en Bali in het Midden van 1847, vol. 1, p. 174.
39 H. Maclaine Pont, "Madjapahit: Poging tot Reconstructie van het Stadsplan, Nagezocht op het Terrein
aan de Hand van den Middeleeuwschen Dichter Prapanca," Oudheidkundig Verslag 1924: 58. See also map
Majapahitsche Restanterikaart in Maclaine Pont, "De Historische Rol van Majapahit: Een Hypothese."
40 Mundardjito, ed., Reneana Induk Arkeologi Bekas Kota Kerajaan Majapahit, p. 166.
41 N. J. Krom, Inleiding tot de Hindoe-Javaansche Kunst, vol. 2 (The Hague: Nijhoff, 1920), p. 106.
42 Wardenaar's father, Willem Wardenaar (1764-1816), was a Dutchman and Counsellor Extraordinary of
the Raad van Indie (Council of the Indies). In 1800, he married his Javanese concubine, Johanna Margaretha
Filius, which consequently allowed him to legitimize their children born previously out of wedlock. One
of these children was Johannes Willem Bartholomeus Wardenaar, born at Semarang in 1785. See De Haan,
"Personalia der Periode van het Engelsch Bestuur over Java 1811-1816," p. 661; F. de Haan, Priangan. De
Preanger-Regentschappen onder het Nederlandsch Bestuur tot 1811, vol. 1 (Batavia/'s-Gravenhage:
Landsdrukkerij, 1910), p. 101; J. G. Taylor, The Social World of Batavia: European and Eurasian in Dutch Asia
(Madison, WI: University of Wisconsin Press, 1983), pp. 95-96; G. H. von Faber, Oud Soerabaia: De
Geschiedenis van Indie's Eerste Koopstad van de Oudste Tijden tot de Instelling van den Gemeenteraad (1906)
(Soerabaia: Gemeente Soerabaia, 1906), p. 123; and J. Hageman J.c.z., Uit het Leven van Wijlen Johan Willem
Bartholomeus Wardenaar (Soerabaya: Kocken, 1869).
43 Gomperts et al., "De Veertiende-Eeuwse Javaanse Hofstad Majapahit," pp. 74—75. Compare Stutterheim,
De Kraton van Majapahit, p. 102.
186 Am rit Gomperts, Arnoud Haag, Peter Carey

The Area to the Northwest of Trowulan

Most of the remains northwest of Trowulan on Wardenaar's plan, A to F, can be
identified as Buddhist. Wardenaar's legend and notes read:

A. "A temple of bricks called Candi Muteran by the inhabitants—according the drawing."
Subtitle of the accompanying drawing in ink: "A decayed temple of bricks in the Forest
of Majapahit called Candi Muteran.
Surveyed on the orders of Mr T.S. Raffles, Lieut.-Governor, on the 5th of October 1815 by
Wardenaar."
B. "A heap of bricks probably the ruins of a temple."

Since Candi Muteran had already vanished by the mid-nineteenth century when Van
Hoevell visited the ruins of Majapahit,44 a copy of Wardenaar's drawing of the
sanctuary is reproduced here (Illustration 3). The candi gives the impression of having
once been a small Buddhist stupa measuring some 2-3 meters wide and 5-8 meters
high. After our initial georeferencing of Wardenaar's plan, we made an on-site search
within a radius of 50 meters of the georeferenced positions of Candi Mutgran and
Wardenaar's "heap of bricks." A woman living nearby pinpointed two spots <A, B>
and stated that a few years ago the soils there had been cleared to a depth of one meter
and the medieval bricks sold. We subsequently identified two spots showing remains
of red brick debris that may have been from the vanished Candi Muteran <A> and the
"heap of bricks" <B> on the satellite image available on Google Earth, which dates
from September 2003. These coincide exactly with the pinpointed spots <A> and <B>,
respectively.
Wardenaar's report then goes on to describe Candi Brahu:

C. "A temple built of bricks called Candi Brahu—as in the drawing."

Pencil note on the back of the accompanying drawing: "The [temple] here shown is
Candi Brahu, situated to the west of Candi Gedung and Candi Tengah, also of bricks.
Very ruined, full of bats, surrounded by forests, big trees."

Candi Brahu <C> was restored in 1990-95, and now has an elevation of 20 meters.45 It
probably represents a large stupa with long stairs oriented to the west.46 Van Hoevell47
states that, according to local tradition, the ashes of the royal family were kept in this
capdi, hence the name "Brahu," which originated from the Javanese word for "ash,"
awu. A Javanese nobleman from the Surakarta kraton visited Trowulan in the
eighteenth century. His account appears in the Modern Javanese text Serat Centini 20.1-
28, dating from 1815. In the text (20.23), the construction of Candi Brahu is said to
resemble that of a mosque.48

44 Van Hoevell, Reis over Jam, Madura en Bali in het Midden van 1847, vol. 1, p. 184.
45 A. R. Kinney, M. J. Klokke, and L. Kieven, Worshipping Siva and Buddha: The Temple Art of East Java
(Honolulu, HI: University of Hawaii Press, 2003), p. 173.
46 A. A. Munandar, Ibukota Majapahit: Masa Jaya dan Pencapaian (Depok: Komunitas Bambu, 2008), p. 101.
47 Van Hoevell, Reis over Java, Madura en Bali in het Midden van 1847, vol. 1, p. 184.
48 Kamajaya, Serat Centhini (Suluk Tambangraras), vol. 1 (Yogyakarta: Yayasan Centhini, 1992), p. 53.
 Mapping Majapahit 187

Wardenaar reports that Candi Gentong had already fallen into decay by 1815:

D. "A temple called Candi Gen tong, entirely ruined."

The georeferenced position on Wardenaar's plan and a local farmer guided us directly
to a structure on site that runs north-south. It consists of the remains of a brick wall
some 10 meters long, which may have once been part of Candi Gentong <D>.
Since the late nineteenth century, the two next candi in Wardenaar's list have both
been named Candi Gentong, but the engineer captain's plan and legend are quite clear
about their original names:

E. "A temple called Candi Gedung as in the drawing."

The accompanying drawing is lost.
F. "A temple called Candi Tengah, entirely fallen."

Candi Gedung and Candi Tengah are still extant remains, now confusingly called
Candi Gentong II <E> and Candi Gentong I <F>, respectively. They both seem to be
Buddhist capdi, possibly representing a mandala-stupa.*9

Jaka Dolog
Jctkct Dolog—literally, the "Plump Lad"—was the local name of an important
historic statue that stands 1.65 meters high. 4950 The statue portrays a Buddhist
Ak$obhya—literally, "The Imperturbable One"—with his right hand touching the
ground in emulation of the Lord Buddha's classic calling-the-earth-to-witness gesture
(Skt. bhumi-sparsa-mudra). "However, the statue lacks all the customary bodily marks
(Skt. laksana) of a Buddha, such as the curly hair, the urna [circle of hair between the
eyebrows] and the cranial protuberance."51 On the orders of the resident of Surabaya,
Adriaan Mauritz Theodore Baron de Salis (1788-1834, in office 1817-22), the statue was
moved to Surabaya and placed in front of the then Residency House at Simpang in
1817, where barren Chinese women came to make offerings to the image to conceive
children <I.2>.52 Wardenaar describes the original location of Jcika Dolog on the site of
the former Majapahit court-capital as follows:

I. "A stone statue of a man— called Jaka Dolog—as in the drawing, which stands on a
small elevation. A few footings in stone for wooden pillars found here suggest that here
was [once] a house or a small temple."
Drawing with note in pencil: "Jaka Dolog in a jati forest on a small elevation [sic]
N[orth]W[est] of the Segaran."_____________________________________________________

49 Edi Triharyantoro and Koos Siti Rochmani, "Candi Gentong: Mandala Stupa Masa Majapahit," Cinandi:
Persembahan Alumni Jurusan Arkeologi Universitas Gadjah Musa Kepada Prof. Dr. H. R. Soekrnono (Yogyakarta:
Panitia Lustrum VII, Jurusan Arkeologi Fakultas Sastra, Universitas Gadjah Mada, 1997), pp. 186-91.
50 See photograph in Gomperts et al., "Rediscovering the Royal Capital of Majapahit," p. 12, figure 1.
51J. Fontein, The Sculpture of Indonesia (Washington, DC: National Gallery of Art, 1990), pp. 54-55.
52 Notulen van het Bataviaasch Genootschap 10 (1872): 143.
188 Am rit Gomperts, Arnoud Haag, Peter Carey

On Wardenaar's plan, the diameter of the circular-shaped elevation measures some 25

to 35 meters. On the basis of the planimetric accuracy of the engineer captain's plan,
the error calculus of our initial georeferencing with other identified points and the
satellite image available at Google Earth, we were able to determine the coordinates of
the original position of Jcikci Dolog <I.0> to within an accuracy of 40 meters. We set out
a circular search area within a radius of 65 meters from the georeferenced position
<I.0>, and subsequently field-walked the search area with a GPS receiver. We were
able to observe a slightly elevated spot some 25 meters long in an north-southern axial
direction, some 20 meters wide and some 0.5-1 meter higher than the surrounding
area, where we found a few old brick fragments on site <I.1>. It is the only elevation—
as well as the only spot—where we observed fragments of brick within the search area.
Hence, point <I.1> was the original site of the Jcikci Dolog statue in October 1815.
Moreover, within the immediate environs (25-50 meters) of Jcikci Dolog's 1815 position,
another recently discovered statue is testimony to this Buddhist environment. A local
villager told us that she watched the excavation, in the early 1980s, of a tall (c. 1.8
meter) statue at point <I.3>, representing—what she described as—a "dewi," or
goddess with bare breasts, standing with a child at her right leg and carrying another
child in her left arm.53 Such a description almost certainly points to the iconography of
Harltl, the Buddhist guardian goddess of children whom parents of prematurely
deceased children worshipped.54 We have gathered more evidence on site, but
limitations of space do not allow us to discuss the full archaeological context of the
Jaka Dolog statue in the present article. Instead, in two separate publications, we argue
that Jaka Dolog represents an image of the legendary figure Bharada, who presided
over the political division of Java in 1052 CE. Moreover, the statue <1.1> stood on the
exact spot of the cemetery known as Awurare/Wurare/Murare/L6mah Tulis/L6mah
Citra, evidence for which is available from newly accessible archaeological data.55

Sites of Menak Jingga, Ratu Cempa, and Bajangratu

In the nineteenth century, the site of Menak Jingga was called Sanggar
Pamalangan.56In Javanese, this denotes an elevated piece of land whose entrance is
barred by cross beams. Wardenaar describes it as follows:

K. "A statue of a woman and a statue of a man sculpted in stone— called Menak Jingga as
seen in drawing—it stands at a larger elevation than the aforementioned statue [i.e., Jaka
Dolog, I], Shards or fragments of black stones sculpted with figurative or flower
motifs—like those belonging to a building—are present here in abundance.
Undoubtedly, the entire elevation consists of similar stones belonging to a collapsed_____

53 Personal communication, Ibu Iklima Nurul Waludin, December 12-13, 2008.

54 See Fontein, The Sculpture of Indonesia, pp. 172-73, 200-1; and Lokesh Chandra, Cultural Horizons of India,
vol. 4, Studies in Tantra and Buddhism, Art and Archaeology, Language and Literature (New Delhi:
International Academy of Indian Culture and Aditya Prakashan, 1995), pp. 176-82.
55 See A. Gomperts, A. Haag, and P. Carey, "Maclaine Pont and the Discovery of Bharada's Hermitage,"
Newsletter of the International Institute for Asian Studies (HAS) 58 (Autumn 2011): 4—5, available online at
http://www.iias.nl/sites/default/ files/ IIAS_NL58_0405.pdf. See further A. Gomperts, A. Haag, and P.
Carey, "The Sage who Divided Java in 1052: Maclaine Pont's Excavation of Mpu Bharada's Hermitage-
Cemetery atLemah Tubs in 1925," Bijdragen tot de Taal-, Lind- en Volkenkunde 168,1 (2012).
56 Van Hoevell, Reis over Java, Madura en Bali in het Midden van 1847, vol. 1, p. 177.
 Mapping Majapahit 189

temple and by excavating this elevation one could surely discover the foundations or
footings of this building. Furthermore, it was surrounded by a moat lined with bricks, of
which a part can still be seen."
Two drawings with a note in pencil: "These two heavily damaged statues are situated on
an elevation in the Majapahit area east of Segaran, surrounded on all four sides by a
moat lined with bricks—beyond doubt a temple of black stones [once] stood here at
what is now the elevation."

The two statues were removed from the site and are kept now at the museum Balai
Penyelamatan, at Trowulan. A lithograph, showing the statues on site in the late 1840s,
is shown in Illustration 4. The statue on the right represents a winged raksasa, 1.41
meters high, holding a small sword on his right hip, and the statue on the left a winged
kinnari, a female celestial musician with her lower body in the shape of a bird.
According to mid-nineteenth-century Trowulan oral tradition, the male and female
figures are depictions of M8nak Jinggct and Dfewi Wahitd.5/ In the Modern Javanese text
Serat Kaynlaning Ringgit Purzva, we read the full story. "Damar Wulan defeats Menak
Jinggci, whose secret protector amulet wesi kuning (literally, "yellow iron"), is betrayed
by his wives Wahita of Bal8ga and Puygngan of Bangkalan." Given the chronological
sequence of the events described, the text can be dated to the fifteenth century.5 758
However, an even earlier date is also possible.
On Wardenaar's 1815 plan, the male and female statues stood at the highest point
of the elevation identified by the present authors as <K.1>. The smaller andesite stones
with animals on the reliefs are still extant, but N. J. Krom states that they were already
incomplete at the beginning of the twentieth century.59The brick-lined moat of the
M8nak Jingga site to which Wardenaar refers has now disappeared. In this context,
Mundardjito mentions the remains of a now vanished covered water conduit
immediately southwest of the elevated area.60 Maclaine Pont saw important brick
remains, measuring 100-150 meters in a north-south axial direction on the eastern side
of Menak Jingga in 1925-26.61 Using these three references, we were able to trace the
course of the 1815 brick-lined moat on a high-resolution satellite image and
subsequently managed to identify three of its former corner points <K.2, K.3, K.4> on
site.62
Wardenaar further describes an important Islamic grave that Trowulan oral
tradition associates with the Islamic wife of one of the last kings of Majapahit:

57J. Rigg, "Tour from Sourabaya, through Kediri, Blitar, Antang, Malang, and Pasuruan, back to
Sourabaya [part 1]," Journal o f the Indian Archipelago 3 (1849): 80-81.
58 Th. Pigeaud, Literature o f Java, Koninklijk Instituut voor Taal-, Land- en Volkenkunde, vol. 2 (Leiden:
Nijhoff, 1968), pp. 361-62.
59 Krom, Inleiding tot de Hindoe-Javaansche Kunst, vol. 2, p. 109.
60 Mundardjito, ed., Rencana Induk Arkeologi Bekas Kota Kerajaan Majapahit, pp. 57, 61.
61 Maclaine Pont, "De Historische Rol van Majapahit: Een Hypothese," see map Majapahitsche
Restantenkaart.
62 The Quickbird satellite image (resolution: 0.6 meter per pixel) was taken on October 2, 2007 (scene
number: 101001000738F301).
190 Am rit Gomperts, Arnoud Haag, Peter Carey

"The large tomb. Here are several tombs. The most important one belongs to Ratu
Cempa; according to the inhabitants, a princess from Sabrang or the opposite coast—and
wife of one of the last kings of Majapahit. The orientation of fhis grave is south and
north—like all graves of Mohammedans. This is why one assumes that this religion was
already known at the time of the last Majapahit kings."

Indeed, in Javanese Islamic graves, the bodies of the deceased are buried in a north-
south direction, with the head to the north and the feet to the south. The gravestone of
Ratu or Putri Cempa <L>, carries the Saka year 1370, which converts into CE 1448.63
Oral tradition tells of a queen originating from Cempa or Champa, the former Muslim
enclave kingdom of the Chams in present-day southern Vietnam. Complementing oral
tradition, Serat Kandaning Ringgit Purwa tells us that her name would have been
Darawati.64 If this oral tradition is correct, she would have been one of the Majapahit
queens or princesses. However, we have not been able to identify her in Noorduyn's
genealogical table of the Majapahit royal family in the fifteenth century.65It should be
noted that Islamic graves and cemeteries in Java are always erected on the boundaries
of settlements, until such time as these expand and the cemeteries have become
engulfed by the surrounding inhabited area. Therefore, it seems that the site of the
grave of Putri Cempa was situated on the northern boundary of the royal capital in
1448.
Wardenaar gives no details of the gate of Bajangratu (see Illustration 5):

H. "A Gate called Gapura Bajangratu, vide drawing."

Nearly all nineteenth-century reports testify that, according to oral tradition, the gate
of Bajangratu had been cursed. Entering it was considered taboo.66The awesome curse
apparently spared the gate from destruction. Bajangratu is a so-called paduraksa, or
roofed gate, with six steps and an eleven-story top <H> rising to a height of 22
meters.6' The orientation of the gate is 18°/198°.68The lintel with holes for door hinges
reveals that the gate opened to the north (see Illustration 6). The reliefs depict scenes
from the Ramayapa and the story of Sri Tanjung mounted on her fish. Similar Sri
Tanjung tales appear on reliefs of Candi Surowono and Candi Jabung, both of which
are fourteenth-century remains. The shape of Bajangratu is quite similar to a gate at
Candi Panataran dating from the fourteenth century.69Therefore, Bajangratu might

63 L. C. Damais, "fitudes Javanaises: 1. Les Tombes Musulmanes datees de Tralaya," Bulletin de VEcole
Frangaise d'Extreme-Orient 48,2 (1957): 390, 407.
64 Pigeaud, Literature of Java, vol. 2, p. 362.
65 J. Noorduyn, "Majapahit in the Fifteenth Century," Bijdragen tot de Taal-, Land- en Voikenkunde 134, 2-3
(1978): 250-51.
66 W. R. van Hoevell, "De Reis over Java, in 1838, van den Gouverneur-Generaal van Nederlandsch Indie,"
21,1-6 (1859): 475; and S. A. Buddingh, Neerlands-Oost-Indie. Reizen over [...] Gedaan Gedurende het Tijdvak
1852-1857, vol. 1 (Rotterdam: Wijt), pp. 324-25.
67 A. J. Bernet Kempers, Ancient Indonesian Art (Amsterdam: C. P. J. van der Peet, 1959), plate 286.
63 Mundardjito, ed., Rencana Induk Arkeologi Bekas Kota Kerajaan Majapahit, p. 97.
69 Kinney et al., Worshipping Simi and Buddha, pp. 170-71, 222, 235; and S. Soejatmi Satari/'Perkiraan
Pertanggalan Gapura Bajang Ratu," Pertemuan Ihniah Arkeologi ke-II, Jakarta, 25-29 Pebruari 1980 (Jakarta:
Proyek Penelitian Purbakala, Departemen P&K, 1982), pp. 241-52.
 Mapping Majapahit 191

also date from the same period and could have been part of a former shrine. There
exist different oral traditions, but they all tell us that the construction of the gate had
not been finished and the residence not inhabited yet at the time of certain events,
which can be dated to the late fifteenth century.70Krom,71 Stutterheim, and Bernet
Kempers72locate Bajangratu just outside and to the east of Prapanca's medieval court-
capital.73On Wardenaar's map, there is no settlement near the gate. At the end of the
nineteenth century, houses appear on the topographical maps. The hamlet
immediately to the south of Bajangratu is called "Kraton," which literally means
"Royal Palace" in Javanese. Such toponyms usually refer to sites where—according to
local oral traditions—a royal residence once existed. Furthermore, the English scholar
of Sundanese linguistics, Jonathan Rigg, who visited Trowulan in the 1840s, adds an
important detail:
On the south side of this building may still be traced, amongst the underwood,
the foundations of walls, as it were of enclosures or courts, but whether they
conducted to a temple or to a great man's dwelling, it is now impossible to say.74
Maclaine Pont observed important remains of adjoining walls adjacent to the east of
the gate, measuring some 200 meters in a north-south axial direction, and parallel
walls 140 meters to the north of the gate measuring some 300 meters east-west. These
have all now vanished.75From these descriptions and the inclining heights adjacent to
the south of the gate, we would conclude that a yard would have been in front of the
gate, with another gate situated to the south of Candi Bajangratu, which led to the
entrance of the shrine.
Wardenaar concludes his legend with this observation:
Furthermore, in the surroundings in the forests there are numerous heaps of
brick pillars, once belonging to houses, walls or temples, beyond doubt there
exist here several antiquities which have not yet been discovered in this
inaccessible jungle.
The forest called Bejijong.
The road from Wir&Sciba [present-day Mojoagung] to Japan [present-day
Mojok8rto].
Surveyed on October, the 7th, 1815 (signed by) Wardenaar.
The incomplete English legend ends: "Surveyed [signed by] C. My. [barely legible]."
We have not been able to identify this individual.

70 Rigg, "Tour from Sourabaya," p. 83; and Knebel, Rapporten van de Commissie in Nederhmdsch-Indie voor
Oudheidkundig Onderzoek op Java en Madoera, pp. 68-69.
71 Krom, Inleiding tot de Hindoe-Javaansche Kunst, vol. 1, p. 112.
72 Bernet Kempers, Ancient Indonesian Art, p. 94.
73 See Gomperts et al., "De Veertiende-Eeuwse Javaanse Hofstad Majapahit alsnog op de Kaart Gezet," p.
74, map 7.
74 Rigg, "Tour from Sourabaya," p. 83.
75 Maclaine Pont, "De Historische Rol van Majapahit: Een Hypothese," see map Majapahitsche
Restantenkaart.
192 Am rit Gomperts, Arnoud Haag, Peter Carey

Finally, the archaeological devastation visited on Trowulan within the area

mapped by Wardenaar has largely involved the removal of subsurface structures. We
estimate that at least 20 million cubic meters of earth with brick remains have been
taken from the area since Wardenaar completed his survey in October 1815.

Conclusion
Wardenaar's plan gives us a unique cartographic image of the Trowulan landscape
on the eve of the colonial exploitation of the area. The quality of his mapping and the
reliability of his report reflect his qualifications as a graduate of the Sgmarang Naval
College and his professional career as an army engineer. Moreover, given that his
linguistic skills allowed him to converse directly with the local Trowulan villagers in
their Javanese mother tongue, one might even argue that the captain engineer was the
first "Indonesian" archaeologist to conduct on-site research at Trowulan. Whatever he
was, he was surely not a Dutchman from The Netherlands who resembled the heads of
the Dutch East Indies Oudheidkundige Dienst, whose self-interested policies defined
the field of Majapahit-Trowulan archaeology in the early to mid-twentieth century.
It is deeply regrettable that Raffles did not publish Wardenaar's map in his History
of Java in 1817. Had he done so, the captain engineer's researches might have set
Trowulan archaeology on a more secure footing for posterity. Even so, the rediscovery
of Wardenaar's 1815 Plan of Majapahit offers us a new benchmark, opening a
perspective on Indonesia's premier archaeological site in an era before the site
destructions of the past two centuries. If the present article has any value, then, may it
contribute to the implementation of a robust regime of archaeological site preservation
so that future generations of archaeologists are not deprived of access to the glory that
was pre-colonial Java.

Appendix: Site Coordinates of Identified Remains

The following list of site coordinates are either positions taken with GPS receivers
(Garmin Foretrex 101 and GPSMAP 60CSx) on site or georeferenced positions of items
shown on Wardenaar's Plan of Majapahit. The accuracies reflect estimated absolute
circular accuracies of GPS readings or those of georeferenced points on maps and
satellite images in relation to the verified ground truth.

W ard en aar's A u th o rs' S ite C o o rd in ates

L eg en d L eg en d U TM W G S 84 z o n e 49M
A A vanished Candi Muteran (see Illustration 3), as 652029 9166246 + 7m
pinpointed to us by a villager
B B vanished pile of bricks, probably the ruins of a 652031 9166215 ± 7 m
temple in 1815, as pinpointed to us by a villager
C C Candi Brahu (triangulation point S.723) 651646 9165991 ± 4 m
D D Candi Gentong, 10 meter-long brick wall 652044 9166009 + 7m
foundations, oriented north-south
E E Candi Gedung, now called Candi Gentong II 652059 9165950 + 7m
F F Candi Tengah, now called Candi Gentong I 652027 9165891± 7 m
 Mapping Majapahit 193

g g G a p u r a Ja tip a s a r , a ls o c a lle d C a n d i 653487 9 1 6 6 0 9 4 + 7 m

W r in g in la w a n g
h h th e r o o fe d g a te o f C a n d i B a ja n g r a tu 654315 9163241 + 7 m
i i .o in itia lly g e o r e fe r e n c e d p o s itio n o f th e s ta tu e o f 651142 9 1 6 4 6 9 6 + 4 0 m
J a k a D o lo g b e fo r e s ite s u r v e y (G o m p e r ts e t a l.,
" D e V e e r tie n d e -E e u w s e Ja v a a n s e H o fs ta d
M a ja p a h it ," p. 77, a p p e n d ix . 2, I)
i .i o n s ite in fe r r e d o r ig in a l p o s itio n o f th e s ta tu e o f 651154 9 1 6 4 6 8 2 + 7 m
J a k a D o lo g (G o m p e r ts e t a l., " T h e S a g e w h o
D iv id e d J a v a in 1 0 5 2 : M a c la in e P o n t's
E x c a v a tio n o f M p u B h a r a d a 's H e r m ita g e -
C e m e te r y a t L e m a h T u lis in 1 9 2 5 ," fig u r e 5)
1.2 c u r r e n t p o s itio n o f th e s ta tu e o f J a k a D o lo g a t 692404 9 1 9 6 5 8 3 + 7 m
T a m a n A p s a r i in S u r a b a y a
1.3 o r ig in a l p o s itio n o f a s ta tu e o f th e B u d d h is t 651141 9 1 6 4 6 6 0 + 7 m
g o d d e s s H a r iti, a s p in p o in te d to u s b y a v illa g e r
k k .i h ig h e s t p o in t a t th e s ite o f M e n a k Jin g g a , th e 652982 9 1 6 4 2 8 9 + 7 m
o r ig in a l lo c a tio n o f th e s ta tu e s s e e n in
Illu s tr a tio n 4
K .2 m o a t o f M e n a k J in g g a site , n o r th e a s te r n c o r n e r 653033 9 1 6 4 3 2 2 + 7 m
K .3 m o a t o f M e n a k J in g g a site , s o u th e a s te r n c o r n e r 653012 9 1 6 4 2 1 7 + 7 m
K .4 m o a t o f M e n a k J in g g a site , s o u th w e s te r n c o r n e r 652909 9164231 + 7 m
l l I s la m ic g r a v e o f R a tu / P u tr i C e m p a w ith th e 652792 9164581 + 10 m
in s c r ip tio n S a k a 1 3 7 0
m m .i ta n k o f S e g a r a n , n o r th e a s te r n c o r n e r 652684 9 1 6 4 5 0 7 + 7 m
M .2 ta n k o f S e g a r a n , s o u th e a s te r n c o r n e r 652620 9 1 6 4 1 3 5 + 7 m
M .3 ta n k o f S e g a r a n , s o u th w e s te r n c o r n e r 652449 9 1 6 4 1 6 7 + 7 m
M .4 ta n k o f S e g a r a n , n o r th w e s te r n c o r n e r 652518 9 1 6 4 5 3 6 + 7 m
n n .i w e ll o f S u m u r W in d u in th e n o r th e a s te r n c o r n e r 652215 9 1 6 2 9 6 4 + 7 m
o f th e S iti In g g il te rra c e
N .2 S iti In g g il te rra c e , n o r th e a s te r n c o r n e r 652232 9 1 6 2 9 7 7 + 7 m
N .3 S iti In g g il te rra c e , s o u th w e s te r n c o r n e r 652150 9 1 6 2 9 1 5 + 7 m
N .4 S iti In g g il te rra c e , p o s itio n o f th e n o w v a n is h e d 652171 9 1 6 2 9 8 6 + 7 m
n o r th w e s te r n c o r n e r , in fe r r e d o n s ite fro m
o b s e r v in g a 9 0 ° a n g le b e tw e e n < N .2 > a n d < N .3 >
w ith a p r is m a tic c o m p a s s
o o .i c e n te r o f th e e le v a tio n w h e r e th e " k i n g 's s m a ll 652402 9 1 6 4 3 6 4 + 7 m
p le a s u r e / g a r d e n h o u s e " o n c e s to o d
0 .2 n o r th e a s te r n s to n e p illa r fo r a tta c h in g a ro y a l 652420 9 1 6 4 4 0 3 + 7 m
e le p h a n t
0 .3 s o u th e a s te r n s to n e p illa r fo r a tta c h in g a r o y a l 652410 9164331 + 7 m
e le p h a n t
0 .4 s p o t o f th e r e c e n tly v a n is h e d s o u th w e s te r n s to n e 652379 9164331 + 7 m
p illa r fo r a tta c h in g a r o y a l e le p h a n t, a s
p in p o in te d to u s b y a v illa g e r
0 .5 n o r th w e s te r n s to n e p illa r fo r a tta c h in g a ro y a l 652398 9 1 6 4 4 0 3 + 7 m
e le p h a n t
p p .i g e o r e fe r e n c e d p o s itio n o f th e P a s s e e r b a a n , 652384 9 1 6 3 9 3 2 + 3 0 m
n a m e ly , th e g r e a t s q u a r e o r a lu n -a lu n
P .2 e x c a v a te d s ite o f S e g a r a n II 652305 9 1 6 4 0 4 0 + 7 m
P .3 e x c a v a te d s ite o f S e g a r a n V 652278 9 1 6 3 8 8 8 + 7 m
P .4 s ite o f b r ic k r e m a in s e x c a v a te d in D e c e m b e r 652306 9 1 6 3 9 8 6 + 7 m
2 0 0 8 , e a s t-w e s t o rie n te d
194 Am rit Gomperts, Arnoud Haag, Peter Carey

Illustration 1. Portrait drawing of J. W. B. Wardenaar, probably dating from the 1860s.

Reprinted from Von Faber, Oud Soembaia, p. 123.

651 0 0 0 * 1 ■ ■ M H H M E M 6 5 2 0 0 0 -. — : 6 5 3 000 w m m m w « * ' i u i i — 6 5 4 0 0 0 ----------------------------------— - 6 5 5 0 0 0 ^ —

; r y
4 w
m

9166000-
F-.V -
*• . M
\ - \ ■*■;_» ' *..
-"V
- C ' :3 ( *
i . n *1 -r V ^ n, j. - • * -• 5 —
*>
\ ” 'i - r~ . r, ; - ’ V | « £
V '* - ^ /* > » ■
• A ^ 2V A * T ~ " '
* >*-%> *
... .. • *y • -,.x — -*-V 1v . »v
*>'• ... . c ; y y . ■
V i-
n M, Sr
______ S =

- ‘ i W h ^^h
'l.>
. K - »x- -h — -
vy >

& - -■ .• a e *
’ -Av yV -tr
WJ5-- . .•»*V ^ | ' - .
'i- ' AV

P S < !‘ - .
* y j p» fe y '
*
, i < y , ' 5 a s *)" V *
_ v; i;
»»*■• ■- Q - 4 V -
9 1 6 4 0 0 0 ------------
« : J. ^ * ' , = ■■ ■■
' "■ , . ,4 • -
v'->» Mf
_ • », * • w *■ » ^
4 ■'* ' **.• V ■ ^
> . jj ...•-' •* ^ 4 *-
Jt 4 ' "/. V - iA *
. \ * .v . * yr" f* * . \ ■*•*»
I- -,v ««* ,* V J
, / > ■■ » t x .
•v - -- ♦N,
»»*

j
r . - A .. j. . »♦• • **
.« .,.. j • * , y
V** JU»»
, . * * “ *•.
j.- ------ ------------- 9 1 6 3 0 0 0 •

N»... ■* >-

T1? -V -*• '* ■*■y fc 'S " / - ’ "

*
A
000 W n - . ~ ~ r ~ - - 6 5 3 0 0 0 r r - ------- . ------- ----------- 6 5 4 o o o 655 s - s H
J T V - 5 651 0 0 0 1

Illustration 2. Plan ofMajapahit, being a copy of Wardenaar's original plan of 1815. The added gridlines
appear at intervals of 1,000 meters and represent coordinates UTM (zone 49 M, WGS84).
© Courtesy of the Trustees of the British Museum.
 Mapping Majapahit 195

Illustration 3. Watercolor on paper of the vanished Candi Muteran <A>. The text of the drawing reads
"fecit Wardenaar." With reference to Verbeek "De Oudheden van Madjapahit in 1815 en 1887," p. 9,
however, the drawing seems to be a copy made in England possibly by the well-known artist William
Daniell (1769-1837). © Courtesy of the Trustees of the British Museum.

Illustration 4. Lithograph entitled "Ruine van Modjo-Paid," by Jan Weissenbruch (1822-90). Reprinted
from C. W. Mieling Javasche Oudheden. Opgedragen mm Z. K. H. Prins Hendrik der Nederlanden (The Hague:
Mieling, 1852-56). See also Bastin and Brommer, Nineteenth Century Prints, p. 70, ill. 44, p. 167, n. 408, 411.
The plate shows the male raksnsn and female kin nan statues in situ at Menak Jingga <K.1> in the late 1840s.
Did the statues face west?
196 Am rit Gomperts, Arnoud Haag, Peter Carey

Illustration 5. Watercolor on paper of the gate of Bajangratu <H>. The text of the drawing reads "fecit
Wardenaar." With reference to Verbeek "De Oudheden van Madjapahit in 1815 en 1887," p. 9, however,
the drawing seems to be a copy made in England, possibly by the well-known artist William Daniell
(1769-1837). © Courtesy of the Trustees of the British Museum.

Illustration 6. Plan of door lintel of the gate of Bajangratu <H> (© Arnoud Haag).