Java程序辅导

 1 
 
 A new introduction to Prambanan:sculpture and dance in ancient Java. A study in 
dance iconography. 
 
Copyright ©  2005 Alessandra Lopez y Royo 
 
My book Prambanan:sculpture and dance in ancient Java.  A study in dance 
iconography was published in 1997  but it referred to research undertaken in the late 
1980s and  the early to mid- 1990s, focused on the dance reliefs of  the Prambanan 
temple complex in Central Java.  Since it came out, I have often been astounded by the 
reactions of readers, when these have been communicated to me.  I have found myself 
quoted as saying what I never intended.   I am not writing this new introduction to the 
book to complain churlishly about those who ‘did not understand me’.  I would like to 
take things further: since the book was published I have had an opportunity to reflect and 
do further research.  This has led me to rethink quite substantially my earlier work and I 
would like to critique a way of thinking which still had me in its grip when I wrote the 
book.  I do not wish, through this introduction, to discard the work done altogether, only  
help the reader to put it in context and make use of the material, bearing in mind these 
changes, in the most appropriate way.  
 
 The dance reliefs in question are those found around the main temple at the 9th century 
CE  Prambanan complex, candi  Siwa (otherwise known as candi  Loro Jonggrang). The 
intense fascination they held for me when I first encountered them was the starting point 
of my earlier research.  Totally obsessed with those images of strong and sinuous bodies 
shown executing elegant movements, I  felt compelled to re-embody their  dancing, 
frozen in sculpture as dance postures .  While attempting to copy those postures, feeling 
them sink in  my own body,  I became aware that they were not random, there was a 
movement logic which determined how they could be strung  together.  To make sense of 
that movement logic I had recourse to chapter 4 of the Natyasastra, a work on drama, 
dance and music which had been extensively referred to in India throughout the 20th 
century restoration of Indian classical dance, especially bharatanatyam,  a process whose 
socio-cultural and political  complexity is discussed by Coorlawala, among others, with 
sensitivity and insight  (Coorlawala 1994).   
 
Chapter 4 of the Natyasastra  discusses  a now obsolete technique, whose foundation is 
the karanas, from a Sanskrit word which literally  means ‘dance action’ and which in this 
instance denotes basic dance movements.   Vatsyayan (1968; 1983a), Bose (1970) and 
Subrahmanyam (1978;1979) have worked on the text discussing  its usefulness as a 
document of older dance  practices and as a classificatory tool – the emphasis on the 
latter  was especially  Vatsyayan’s (1968, 1983, 1983a) .  Subrahmanyam in particular, 
being a practitioner,  set about  reconstructing and re-embodying  the technique that is 
described in the text (Subrahmanyam 1978; also see Balakrishnan 1992), referring 
simultaneously  to Indian temple sculpture and iconographic representation of these 
 2 
karanas1.     Meanwhile, the Javanese scholar Edi Sedyawati was working on dance in the 
context of an assessment of the Indo-Indonesian relationship of ancient times.  She had 
begun to identify the movements frozen in the Prambanan dance reliefs as sub-units of 
karanas (caris and sthanas and hastas, that is, leg movements, stances/postures and hand 
gestures) (Sedyawati 1982).  
 
Working within parameters already set by others, I have argued for an identification of 
the dance reliefs as representations of karanas of the Natyasastra and provided specific 
correlations between individual  karanas listed in the text and each relief,  situating  the 
analysis in the context of the debate around the issue of Indianisation and localisation of 
Indian influences in Southeast Asia,  the ‘India –Java axis’,  as Wright calls it (Wright 
2002,175).  
 
Out of that endeavour the most important  issues to emerge  - and which need to be 
reviewed  - are  the ‘old ideas’, those which constitute  the terms of an old discourse:   the 
India – Java  axis, the  Indian versus Javanese terminology to describe the dance 
movements of  the Prambanan reliefs, the role of texts such as the ‘foreign’ Natyasastra 
and  the ‘indigenous’ Nawanatya for understanding  Javanese dance,  and finally,  the use 
of Labanalysis and notation for analysis and recording. 
 
The dance reliefs: dislodging the ‘old ideas’ 
The following discussion focuses on the body, body movements and how these are  
analysed, recorded and documented in a cross-cultural  context. This brings awareness  of 
the many issues surrounding  the moving and dancing body  in the present and in the past.   
Movement classification 
 I have been taken to task for not  using  a Javanese system of movement classification.  
The reason for not doing so is that  there is not any that can be applied and/or adapted to 
dance movements as early as those of the Prambanan reliefs.  The dance forms and 
genres current in Java today do not cover the same range of movements and it is only in 
more recent years that they have been consciously organised in terms of ‘technique’, not 
without a substantial debt to both Labanalysis and the  Natyasastra, as will be seen –  I 
use here the  term technique in a very specific sense.  This more recent process of 
systematising Javanese and Balinese dance movements has been stimulated by an 
encounter with Euro-American  modes of dance analysis.  The lack of an extensive   
Javanese dance terminology co-eval with the reliefs and  the lack of any codification of 
dance movements in a written form  until the 19th and 20th century  could perhaps be 
taken   as an indication that the practice of classifying dance movements was  not  a 
favourite Javanese pursuit, rather than believing that there must have been one which  
was subsequently  lost – though of course everything is possible and I am not ruling out t 
that there may have been manuals, though we have none extant.  The Javanese texts 
written in the 19th and early 20th century, with an account of how to perform specific 
movements  of Javanese court dancing may  have been prompted  by an engagement with 
modernity, in a postcolonial mode,  as Brakel points out (Brakel – Papenhuyzen 1993; 
1995).  Why did I not use these indigenous texts?  I felt that though complex and rich in 
                                                 
1 Subrahmanyam was my mentor at the beginning of my   re-embodiment process, back in 1993.   
 3 
terminology, their model was Javanese court dancing and they were thus fit to describe 
only Javanese court dancing.  Also, as Brakel notes, lack of standardisation in the 
manuscripts of the Surakarta tradition causes problems with the interpretation of 
terminology (Brakel-Papenhuyzen 1995,ix). 
 
  Dance academies in Java and Bali, such as STSI and ISI2,  are  keen to  classify and 
analyse movements because it suits a way of teaching which has inevitably changed with 
the introduction of the academies themselves –  with large classes, rather than one to one 
tuition and  compulsory study of dance history and theory and because of the need to 
improve documentation  and recording of  dance, without having to  depend exclusively 
on personal memory and recollection3.  A number of projects whose purpose was to 
classify the movements of Javanese and Balinese dance were sponsored by the 
Indonesian Department of Culture in the 1970s and throughout the 1980s and 1990s, 
some of which were initiated by Edi Sedyawati who for some years held the post of 
Director of Culture4.   Dissertations have been written by STSI graduates in which they 
have attempted an analysis of specific dances and dance genres, using these recently 
created systems.  The academies are also keen on the use of Labanotation, strongly 
encouraged by distinguished performer/academics such as Soedarsono and Madé 
Bandem, both former rectors of ISI, Yogyakarta  and STSI Denpasar respectively.  
Soedarsono has  used it in his writings (Soedarsono 1990). 
 
Between 1999 and 2002 I had the opportunity to work in greater depth with the dance 
movements of the Prambanan reliefs.  I worked on projects5 where I was involved in 
using, at times more, at times less successfully, new technologies for an analysis of the 
movements in the reliefs.  But most importantly, the projects gave me scope to  work on 
issues of re-embodiment with Javanese, Indian and British  contemporary dancers6.  I also 
began to engage in further dance analysis together with Ni Madé Pujawati, a Balinese 
dancer with whom I have been working on issues of Balinese dance technique with 
reference to lègong dance (a genre of Balinese dance) through practice-based research.  
All this work has been significant and has allowed me to reassess the Natyasastra, to 
refine concepts of dance technique and to reformulate my work with the karanas.  
 
  I began to read the Natyasastra interpreting it as a classificatory tool, following in the 
steps of Vatsyayan’s very lucid analyses of this text.  I ceased to regard it  exclusively as 
                                                 
2 STSI is short for Sekolah Tinggi Seni Indonesia and ISI is Institut Seni Indonesia.   
3 I am not here  implying that  embodied memory should be dismissed, I am only saying that there is a need 
to complement it with other ways of recording. 
4 Vatsyayan  acted as mentor to Edi Sedyawati in her earlier explorations of the Natyasastra in relation to 
the dance reliefs of Prambanan (Vatsyayan pers.com). This is very significant, in view of Sedyawati’s later 
engagement with the above mentioned projects which aimed at classifying the movements of Javanese and 
Balinese dancing, following a complex system of body movement  analysis articulated through verbal 
description.  The movement categories correspond to those of the Natyasastra, though the nomenclature is 
taken from the dance forms themselves. 
5  Funding for all these projects was gratefully received from  the Arts and Humanities Research Board and  
the Getty Research Program.  I should also thank the University of Surrey, the University of Oxford and 
Roehampton University for giving me an academic home during this research. 
6 Didik Bambang Wahyudi, Mugiyono, Vena Ramphal and Liz Lea. I am very grateful to them for their 
input and also to Jean Johnson Jones, University of Surrey. 
 4 
a dance historical text. This shift in thinking meant that the actual date of the 
Natyasastra7 and the issue of whether it was ever really known in ancient Java as a 
written text were no longer central.  What was crucial was the paradigmatic concept of 
technique   retrievable  from  the Indian text.  The Natyasastra  divides the human body 
into planes and breaks up the human body into smaller parts,  segmenting the bodily 
actions,  which are   then reassembled and co-ordinated into a whole. Thus a study of the  
Natyasastra specifically in Vatsyayan’s analysis of Indian classical dance, reveals a      
“notion of technique …that functions  as  a framework that   incorporates a set of criteria 
for extracting the characteristic features of a system of stylised bodily actions” (O’Shea 
2000,82) and the way these are systematically joined.  This concept of technique 
retrievable from the Natyasastra differs considerably from the way technique is 
conceptualised in western dance forms, as O’Shea notes (O’Shea 2000, 82), where   
technique refers to a set of movement exercises and dance steps practised in a class 
situation but it does not necessarily entail a complex classification of body movement as 
put forward by Vatsyayan, through her reading of the Sanskrit textual material.8 
 
The Natyasastra classifications and codifications of movement thus provide an analytical 
tool, a model, which has the potential to be applied  cross-culturally.  This is  because the 
technique can be assembled from the dance actions of each  dance and then slotted into a 
category which doubles up as dance movement unit, susceptible to subdivision –  a dance 
karana.  In practice, I have used this model for a theory of corporeality of ancient 
Javanese dancing,  organising  the analysis around the concept of dance technique, 
centred on the dance karana.  By arranging the movements into karanas,  I have in fact 
used a tool which emerges from the dance itself - here I specifically mean the  dance 
movements of the Prambanan reliefs. This occurred as the dance was being re-embodied: 
the re-embodiment was led by  the segmentation process and subsequent co-ordination of 
the limbs’ movements9.   That the analytical tool should emerge from the dance itself was 
important: I did not wish to superimpose analytical categories  external to the dance, as 
would have been the case if I had used contemporary Javanese court dancing  as a 
paradigm.  
 
 In other words, stimulated by Vatsyayan’s reading of the text,  I used the Natyasastra  
categories and classifications instead of more commonly relied upon  systems such as  
Laban Movement Analysis (LMA).   O’Shea notes that the understanding of dance as a 
conscious cultural practice, enabled through Vatsyayan’s reading of the Natyasastra and 
similar texts   
 
                                                 
7 The Natyasastra is reputed to be not later than the 5th century CE (cf. Vatsyayan 19682,70: text and Gupta 
sculpture seem to be  closely linked). 
8 It should be noted however that dance teachers aware of Labanalysis may apply analitycal concepts drawn 
from Laban   in the studio while teaching.  Hutchinson –Guest writes: “ each exercise is carefully taken 
apart and the spatial pattern, use of the body and the involvement of various muscles (the anatomical facts) 
are discussed, discovered in oneself and observed in others” (Hutchinson-Guest 1984,160) 
 
9 Dancer Sandhya Purecha has worked  in depth on this  issue of segmentation.as a prerequisite for re-
embodiment.  Purecha presented a lecture demonstration at the Nehru Centre in London on 26th October 
2002 discussing her research work  on the text Abhinaya Darpana.  
 5 
 can be described verbally through its component features. (emphasis mine). This 
 view in turn provides a way into articulating divergences and similarities in the 
 construction of corporeality among different cultures or time 
 periods…Vatsyayan’s paradigm also facilitates the interpretation of the cultural 
 priorities evident in a dance form while not reducing dance to the status of 
 reflecting its culture , time period or social group.  This methodological 
 framework situates dance as an active cultural participant in relation to other 
 systems of thought (O’Shea 2000,84).   
 
 Labanalysis and Labanotation  versus verbal description 
The use of LMA  (and its attendant notation system, Labanotation) is  not unproblematic  
- hence my reluctance to privilege this over other systems.10 Labanalysis and 
Labanotation, on a par with other western  notation systems, proceed on the assumption  
that analysis of movement should be anatomically based to guarantee objectivity11.  This 
is  not necessarily applicable in the same way to all cultural contexts. In western dance 
for example the body is viewed in terms of the lines of its musculature but elsewhere, as 
in India, the dancing body is constructed around the articulation of joints (in relation to 
one another), and in terms of alignment and medians (Vatsyayan 1983a; O’Shea 2000). 
Thus theorising the moving, dancing body from an anatomical standpoint as understood 
in the West may be unhelpful and cause a distortion of perception when applied to non-
western dance.  Would the methodology derived from the Natyasastra not be equally 
alien?   Not when used to work with dance forms which are compatible, in other words, 
which seem to give emphasis to  joint articulation and alignment – without implying that 
one has to postulate  a derivation of such forms from ancient Indian dancing.    For 
example, Ketut Rota, formerly a student at STSI Denpasar (now ISI), has used  the 
classificatory principles of the Natyasastra as a reference point in her analysis of lègong 
dancing, using the Sanskrit  text (through the interpretation given to it by Naidu, Naidu 
and Pantulu) (Ketut Rota 1991; Naidu and Pantulu 1936)12.   
                                                 
10 It should be clarified at this point that though I have some reservations on the  use of  LMA as the only 
method of movement analysis, I am also very aware of its  enormous usefulness.  Dance documentation 
could not have developed the way it has without the effort of Laban analysts and Labanotators and it would 
be  an injustice  not   to acknowledge their very positive input.  
11 However,  the anatomical basis of Labanalysis does not reach the depth of bio-mechanical analysis. 
Biomechanics is the science which analyses the mechanical characteristics of movement. It is through a 
biomechanical analysis that we can gain an understanding of the mechanisms through which  movement 
was achieved and allow for  comparisons of movement. Biomechanical analyses have been used for clinical 
and non-clinical analyses ranging from the assessment of amputee gait to the development of new 
gymnastic techniques. The source data for biomechanical analysis is often digitised motion data.  Since the 
flicker of the eye is 10-12 frames per second, it cannot see particularly fast motion and the motion is lost as 
soon as it has happened. These factors led early bio-mechanists to develop motion capture systems which 
could record movement at higher frequencies and produce permanent copies. With the advancement of 
video and computer technology, motion systems operating at frequencies of up to 240Hz can be used to 
analyse three-dimensional movements as a matter of course . As biomechanics can characterise the 
movement,  describe and quantitatively analyse it , were it applied to dance  it would enable an 
understanding of  the fundamental movement pattern of a dance.  I am grateful to bio-mechanist Siobhan 
Strike, School of life Sciences, Roehampton University,  for  demonstrating  the complexity of bio-
mechanical analysis through a  motion capture session. 
12 O’Shea is of the opinion  that Vatsyayan’s model of dance analysis owes as much to the Natyasastra and 
the Sanskrit textual tradition as to her experience of Labanalysis: “By linking this interpretation with the 
 6 
 
There is also, as mentioned,  another interesting feature in carrying out an  analysis  using 
the Natyasastra’s  conceptualisation of the dancing body: the recourse to and 
enhancement of verbal description to analyse a dance movement, following a  detailed  
system of classification.  This  raises a host of new questions.  Most western notators and 
movement analysts have tended to regard verbal description   as the worst possible way 
to approach  an analysis:  it is regarded as  lacking  objectivity and  as an impediment to  
international communication because it relies on language (Hutchinson –Guest 1984,10).  
The implication is always that of a superiority of  a (Laban)  score which  would  not 
have  similar shortcomings.  Ann Hutchinson Guest  in a  section entitled  “Why not word 
descriptions?” (Hutchinson Guest 1984, 2-14) gives   quite a number of examples which 
are meant to illustrate the ambiguity of language in recording movement; but  the 
examples are ambiguous   because the description is laconic in the extreme.  More words 
and the use of imagery would have made the verbal description  less ambiguous than we 
are led to believe.    
 
Yet,   she concludes by saying that “though words are not the tool with which to record 
movement, they are a vital means of communication in learning to understand and master 
movement…To be considered is whether a notation system should only state the practical 
facts of a movement  or should also include the idea behind it, the concept, the motivation 
[to be expressed in words]”. (Hutchinson-Guest 1984,14).  In Hutchinson Guest’s view  
the notation system deals  with ‘facts’,  therefore  its objectivity remains  unquestioned13. 
Karanas as a way of understanding dance 
Viewing karanas and their sub-units as analytical tools is a departure from what I argued 
in 1997.  It was only at a later stage that I began to formulate  this  more  fitting 
conceptualisation  for what I had learnt through my body and had attempted to explicate 
in that earlier phase of my work. I began the process of articulating my reflections only in 
later papers ( Iyer1999; Lopez y Royo 2001; 2004). This, coupled  with further  practice-
based research,  led to a major shift in my thinking. In the 1997 work  I  relied on the 
categories of the Natyasastra  as indicating and representing an obsolete technique.  I 
took them at face value  and did not fully explore them as epistemology, that is, a  system  
for  organising  body movements and  conceptualisations of body14.  Though in practice I 
                                                                                                                                                 
Laban analyst’s ability to observe movement vocabularies through their component parts, Vatsyayan 
develops a model of dance analysis applicable to movement forms beyond those referred to by the Sanskrit  
aesthetic theory texts” (O’Shea 2000, 83).  I would however be wary of seeing a possible amalgam of the 
two systems. 
13  Even more significant, and highly contradictory,  is the analogy with language that  Hutchinson Guest is  
keen to pursue when discussing  dance in general.  This  is seen as “a language of expressive gestures 
through which non-verbal communication can be achieved. Like verbal language it has basic ‘parts of 
speech’” [such as nouns, verbs, adverbs] (Hutchinson Guest 1970,17).   
 
14 I think it is fitting to be reminded  here of what  Hughes-Freeland wrote in 1991 “ It is important to 
recognise that the studies of Indian culture and thought have yielded important findings for the study of 
culture and society in general , not on a diffusionist basis but on the grounds of discovering non-western  
philosophical systems which have important implications for classificatory  systems and concepts in non-
western societies” (Hughes-Freeland 1991,147, emphasis mine) 
 
 7 
used the karanas  as movement categories, I did not dwell on  the theoretical implications 
of using them as such.15   
 
  Then I understood karanas principally as an ancient dance technique16 which seemed, 
from the evidence I had gathered,  to have been known in India and in Southeast Asia.  
Indeed,  the original karanas of the Natyasastra  were based on dance movements which 
would have been known through practice – here I agree with most  scholars and   
commentators  of Indian dance who maintain that the Natyasastra took stock of what was 
current practice in India at the time of its writing and set an aesthetic  standard for  
performance.   What I overlooked, however,  was that  as they were codified, the karanas   
ceased, after some time,  to belong to any specific  dance form and style: they became a 
dance movement category.  This  turned them into  a classificatory tool and a system for 
movement analysis through its codification - Vatsyayan clearly articulates  this in her 
study of the Sarangapani temple dance sculptures (Vatsyayan 1983b). 
   
One can thus argue, and indeed I now do,  that using the karanas to analyse the dance 
reliefs of Prambanan can  be totally divorced from the question of Indianness and 
Javaneseness of the dance, because they are an analytical tool . Today I would argue  that 
since karanas  are analytical categories of dance movements,  to ask whether the 
Prambanan dance movements, which can be identified as dance units  with the help of 
codified  karanas,  are Indian or Javanese is meaningless.    They are found on a Javanese 
temple, thus it is reasonable to see them as Javanese – made by Javanese craftsmen and 
representing Javanese dance.   Working with the Natyasastra in connection with the 
dance reliefs of Prambanan is no more than a way to make sense of the movement 
articulation of the dance captured in those reliefs: the Natyasastra enables a viable 
interpretation.  
The’past as wished for’ 
As I continued to engage with the dance reliefs, I came to realise that in my earlier 
exposition, I had used a predictable and inadequate  frame to view the relationship of the 
Prambanan dance reliefs with  the karanas of the Natyasastra – the relationship of 
ancient  Indian and Javanese culture.  With an area studies background, this was 
somewhat inevitable.  
 
The debate on the cultural interaction between ancient India and Java, enticing though it 
may be, goes round and round in circles, in a fix to find ‘origin’.  It is a  political issue 
and it has  to do with the present and how  the past is interpreted to suit it. The theory of 
‘Indianization’ of Indonesia and Southeast Asia served a purpose during the colonial era, 
just as did the ‘localisation of Indian influence’ in post-independence Indonesia.  So did 
my attempt to see ‘recycling’ as a possibility, sprung from a desire to look for 
                                                 
15 To complicate things further, I did not   translate the Sanskrit nomenclature  into terminology which 
could be more easily understood by contemporary readers, not accustomed to Sanskrit – this made the 
analysis quite inaccessible. 
 
16 Let it be clear however that I never took them for  a dance form or genre, as is often  assumed, 
confusingly,  by my reviewers. 
 
 8 
alternatives, and  influenced by notions of cosmopolitanism, particularly in the sense of 
the Sankrit cosmopolis theorised by Pollock (1996)17 .   As Shanks explains “We all 
produce a ‘past  as wished for’ in the sense of pasts which we find empirically, 
theoretically, aesthetically and archaeologically ‘satisfying’” (Shanks 1997,166).  
 
Whereas in 1997 I felt obliged to engage with the issue of the relationship between India 
and Java, trying to give pointers for  further lines of enquiry to change entrenched 
attitudes towards and perceptions of this relationship,  I now feel it was unwise of me to 
rely on such a frame for my discussion.   Somewhat naively, I thought that by engaging 
with the issue I could stimulate  some further reflections on how the past is constructed 
for present political needs.    I never expected  that,  far from encouraging reflexivity,  the 
debate  would not go  further than considering  the issue of  Indian versus Javanese 
dance.  In assigning an Indian identity to the dance movements of the Prambanan  reliefs  
because of an attempt at  making  sense of the movements using Indian dance 
scholarship,  it seems to me that one is  being influenced  by some misconception about   
ethnicity,  here “predicated on a conjunction of specific material culture objects and an 
essential notion of identity” (Shanks 1997, 174).  It had genuinely not occurred to me that 
the dance of the Prambanan reliefs could be anything but Javanese. 
Red herrings and wild goose chases 
Connected with, but also distinct from the issue of the India-Java axis, is the text 
Nawanatya, which chapter 4 of the book discusses.  The text  has held tremendous 
fascination over a great number of scholars of ancient Java, partly because Pigeaud made 
it relatively popular in his translation of excerpts from it (1963).  It has been thought  to 
be variously connected with the Natyasastra and to be particularly valuable for an 
understanding of ancient Javanese performance.   I have myself thought this was so.  I 
wrote more about Nawanatya  three years after the Prambanan was published (Iyer 
2000).  But this was an essay full of hurried conclusions.  Because Pigeaud used it for his 
monumental Java in the 14th century   it has now become axiomatic that this is a Javanese 
Majapahit text and I took it to be so.  But in 200218 I had the opportunity to do some 
further research about this work in Bali,  and have now come to very different 
conclusions.  There is no evidence to suggest that Nawanatya is a Javanese Majapahit 
work nor is it as old as is maintained by some scholars  –  reading it together with Old 
Javanese scholar Kuntara Wiryamartana19 made me aware that  indeed the language of 
this text is fairly recent. The Majapahit dating is only an assumption, based on a previous 
authoritative reading and on the proposition that the court terminology refers to 
                                                 
17   One could   see here a ‘past as wished for’ influenced by contemporary ideas of global networks and the 
spread of English as an international lingua franca  
 
18 I am grateful  the British Academy  Committee for South East Asian Studies for their award which 
enabled the fieldtrip. 
19 Dr Wiryamartana was at Oxford as a visiting fellow at All Souls College throughout  the spring of 2003  
 
and I was able to work on the text under his guidance during that period . 
 
 9 
Majapahit Java.  On reflection,  this is not a strong argument; Nawanatya could on the 
contrary  be a late work recast in the language of Majapahit to lend it greater authority20.   
 
The text is found only in East Bali and there is no reason to doubt that it was a Balinese 
text composed in Bali/Lombok.  There are very few copies in Bali of the original 
manuscript and the only ones available are all from the  Karangasem regency; there are 
no  copies or similar  works in Java21.     Nawanatya is  a book of  behaviour – oddly 
resembling  those court manuals from  Italy, such as the Book of the Courtier by 
Baldassar Castiglione.    As such, it  presents various sets of rules of conduct  through 
which  the perfect male courtier (the servant) is moulded.  Disciplinary practices (diet,  
fitness, adornment, posture, mental attitudes) are highlighted which produce the bodies of 
their concern; primarily male bodies.  The female body and  the female presence is 
marginalised, but  it is not absent and it  can be retrieved from a reading of the text. 
 
 All the links that have been drawn between Nawanatya and the Natyasastra are very 
tenuous and  unsuitably  framed.  Nawanatya does not in any way provide a classification 
of dance movements.  Its performance vocabulary has a Sanskritic component, being in 
Old Javanese,  but this in itself cannot help to identify movement patterns nor indeed 
performance genres – though in terms of the India –Java – Bali axis it could be  read  as 
providing evidence of some prior knowledge of the tradition of the Natyasastra, which 
too engages in discussions of courtly behaviour22.  
 
Nawanatya is a significant text for performance, but for this kind of reading we need to 
have recourse to  a  theoretical frame which differs substantially from that employed by 
earlier researchers, including myself.  Seen in this new light, Nawanatya ceases to have 
any significance in terms of the old question of how it relates to Indian material. The 
content of  Nawanatya   directly impinges on performance practice – Balinese 
performance in particular. The world evoked by the Nawanatya  ‘actors’  is close to the 
world of Gambuh performance – Formaggia includes the text  among the manuscripts 
referred to in compiling  her collection of essays  on Gambuh    (Formaggia 2000,360 ) - 
and  relevant  to the world of baris and lègong dancing.  This is not a manual prescribing 
or describing  steps and giving specific stage directions – but the codification of 
behaviour and conduct one finds in it establishes a new connection with Balinese living 
performance practices. 
                                                 
20 Helen Creese makes a similar point when she writes:” the Balinese courts of Lombok in the mid-
eighteenth century …were given Javanese place names. The creation of new courts was also marked by the 
relocation and recopying of texts designed to encapsulate the earlier moral, social religious and political 
universe shared with Bali and Java.  For this reason the elaborate descriptions of Javanese landscapes and 
courts in the texts, had their very counterparts in local communities both in Bali itself and in Lombok” 
(Creese 2000, 32-33) 
 
21 However, a  fitting term of comparison is the 1925 text from the Yogyakarta court SÏrat Kridhwayangga by 
Sastrakartika,  mentioned by Brakel-Papenhuyzen.  In this work, the characters of wayang wong theatre are 
connected with the behaviour to be displayed in relation to the ruler by specific individuals or groups (1993,68) 
22 One should also add here the vast Niti (politics, rule) literature available in Old Javanese and Balinese. 
 10 
  
 The Prambanan dance units as  open works 
I would like at this point to leave aside the issue of Indianness and Javaneseness of the 
dance in the Prambanan reliefs  and retrace my steps.  Segmenting, analysing, 
reconstituting the dance movements of the Prambanan reliefs using karanas  for this  
process (I could have used LMA or any other system, I chose the Natyasastra  because it 
seemed to be  more appropriate) cannot be done without  embodiment.   Here, theory 
meets practice and the two  mutually sustain  each other.     But are the re-embodied 
movements ‘real’ karanas ?  There are  two ways of considering this question. The first is 
to understand it  as effectively saying:  though  karanas  are a classificatory system, do  
they exist as dance units in their own right?  
   
I have already suggested that when karanas  were codified, this turned them into 
abstractions – they ceased to have a connection with a specific dance form or a style of 
execution.  The main problem here is one of language, in more than one sense.  In  the 
case of the  dance of the  Prambanan reliefs, due to a lack of technical terminology 
pertaining to the dance, the classification has taken over and doubled up as the  
movement nomenclature of the technique  analysed.  This would not happen in the case 
of dance forms with their own technical vocabulary – though they can be analysed in 
terms of angas, upangas, caris , karanas  etc (the categories of the Natyasastra)  the 
original nomenclature of their movement would not be altered. The argument can only be  
barely sketched here;  there are indeed important philosophical implications concerning 
nomenclature, classification and the construction of meaning and they should  be further 
examined,  but in the context of this introduction I shall have to withhold their discussion.  
 
If, however, those who ask about ‘real’ karanas  mean ‘the original karanas’, then we are  
on shaky grounds.   The dance reliefs can be analysed through karanas.  But in the act of 
re-embodying them they come to life as dance movements in their own right and they are 
reconstituted as a series of small dance units, which  can be danced in whatever 
combination, undergoing,  if so desired, further fragmentation and reassemblage, fusion 
and fission.   They exist in the bodies of those who have gone through the process of 
reconstituting  them  and can be taught, passed on from one body to another through  
demonstration and training.  What  is a ‘real’ karana ?   Are we  expecting  some 
guarantee of  absolute authenticity?  Of truth?   Shanks and Tilley have eloquently 
written : 
   The truth of the past is metaphorical.  It is to be found in the traces of the past, it 
 is present in-itself in the past, present with us.  At the same time the traces of the 
 past point to an absent truth, a truth outside the past found in the reception of the 
 traces of the interpreting archaeologist…So we do not begin with the truth of the 
 past, produced by people in the past, and end with that truth revealed by the 
 archaeologist in the archaeological text.  We find our affinity with the past 
 through our difference to it, through practice which links past and present.  Truth 
 is delivered by the interpreting archaeologist on a detour away from the past,  a 
 detour to truth (Shanks and Tilley 1992,20).23 
                                                 
23 Again,  this is an issue I did not clarify in 1997.  But my further work with dancers made me reflect on it.     
I wrote elsewhere that though karanas may have been performed in India or Thailand or Java, the quality of 
 11 
 
The re-embodied dance movements of the Prambanan reliefs  can have a new lease of life 
as a contemporary dance technique.  The choreographic possibilities held by re-
embodying the dance units  of  the reliefs are endless and the performance rules are 
entirely of one’s choice: through a process of segmentation,  reassemblage and co-
ordination a new set of movements has been re-embodied, reconstituting what to us today 
are dance units. In saying this, I am pointing out that the dance units retrievable from 
Prambanan may not have been regarded as dance units back in the 9th century:  we do not 
know how the technique was conceived of at that point in time and  where its caesurae 
were.  But they can be reconstituted as dance phrases which to us become dance units for 
working on choreographies.   
 
 Issues of movement aesthetics, composition , repertoire are  open-ended, to be 
determined and decided by the individual dancer/choreographer who works  with the 
reconstituted material.  Having engaged  with  Didik Bambang Wahyudi and  Mugiyono 
Kasido  (and with my own body) in  re-embodying the dance units of Prambanan, helped 
me to re-evaluate  the issue of assigning a quality to  the movements: dancing them is  a 
performative  act, in which people – dancers and spectators – negotiate conventions.   
 
Unlike an established dance technique, with a fixed style of execution imparted  as a 
model by the teacher, the dance units of Prambanan are minimally pre-defined. They no 
longer exist solely as an abstraction but they are fluid and open to interpretation.    
Thinking in terms of 9th century Siwa Tandawa dances  springing  to life would  be 
entirely off the point.  Rather, it makes  sense to think of the Prambanan dance units as 
mini-open works,  non-representational  – each   dance unit being equivalent to an open 
work, in the way Rubidge conceives of it, but reducing it to a smaller scale.     Thus there 
is   “ no original author-determined performance or instantiation against which all other 
performances can be compared …The open work has no original, nor has it an ideal 
performance.  It has instead clear originating ideas from which performances are 
generated. It is in these ideas, not in the external forms, that the identity of the open work 
lies” (Rubidge 2000,212-213) 
 
In conclusion… 
 The misgivings that my earlier work on the dance reliefs of Prambanan seems to have 
caused and  my rethinking  of this  work have compelled me to reassess it.  My 
theoretical frame and my conclusions have  changed, in the light of further research and a 
self-reflexive engagement  with it.  The Natyasastra, which I have used to interpret the 
dance reliefs, is now to be understood  as a  classificatory and analytical tool, rather than 
                                                                                                                                                 
movement and the way the karanas  were strung together in choreographic compositions would have been 
very different and it was this which would have given them an Indian, Thai or Javanese identity  (Iyer1999; 
Lopez y Royo 2001).  When I wrote this   I was trying to make sense of the palpable difference in 
execution that can be noted  when  the same movement pattern is danced by different bodies.  Then I 
understood karanas as ‘real’ and as having been spread from India to Southeast Asia, and  I saw them as 
surviving in living performance practices in a number of Southeast Asian dance genres.   I had not yet 
arrived at  the conclusion that   karanas  are principally a classificatory and codification system. It is only 
in the act of re-embodying the movements, whose analysis and reconstitution  the karanas enable,  that they 
can  be danced, with a new  identity which has to be  negotiated. 
 12 
a historical document  inscribed in the India-Java axis debate.  As the analysis of the 
reliefs proceeds, the dance –units reconstituted through the analysis are re-embodied and 
the re-embodiment itself feeds into the analysis.  
 
How is then the book to be used? The analysis of the dance reliefs using definitions of 
cari, hasta, sthana; the segmentation of their movements;  the re-embodiment which 
relies on  the principle of the shortest route from one position to another, chart out a 
methodology.  The identification with the 108 karanas of the Natyasatra is an attempt to 
slot the reconstituted units into movement categories, represented by each karana, each a 
dance unit.  It is first and foremost an interpretation and as such it is open ended. The re-
embodiment turns the newly reconstituted dance units from abstract categories into mini 
open works: the dancer/choreographer dances them on her own terms, supplying 
interpretative nuancing.  The Prambanan site is thus danced into the present:  an 
embodied archaeology.  
 
 
 
 
 
 
 13 
References 
Bose, Mandakranta 
1970 Classical Indian Dancing: a Glossary  Calcutta: General Printers and 
Publishers 
Brakel-Papenhuyzen, Clara 
1992 Of sastra, penget and pratelan: the development of Javanese dance notation. 
Bijdragen tot de Taal-, Land-  en Volkenkunde 148: 3-21 
1995 Classical Javanese Dance Leiden: KITLV Press 
Coorlawala, Uttara Asha  
1994 Classical and Contemporary Indian dance. Overview, Criteria and 
Choreographic Analysis. Unpublished PhD dissertation, New York University 
Creese, Helen 
2000 In search of Majapahit: the transformation of Balinese identity in To Change 
Bali.  Essays in Honour of I Gusti Ngurah Bagus. 15-46 eds Adrian Vickers,  I 
Nyoman Darma Putra and Michele Ford, Wollongong: Institue of Social Change 
and Critical Inquiry, University of Wollongong/Denpasar: Bali Post 
Formaggia, Maria Cristina 
 2000 Gambuh Drama Tari Bali Jakarta: Yayasan Lontar 
Hughes-Freeland, Felicia 
1991 Javanese visual performance and the Indian mystique in The art and culture 
of Southeast Asia : 125-150 ed. Lokesh Chandra New Delhi: Aditya Prakashan 
Hutchinson Guest, Ann 
1970 Labanotation. The system of analysing and recording movement. London: 
Dance Books and New York: Theatre Arts Books. [1st edn 1954] 
1984. Dance Notation The process of recording movement on paper.London: 
Dance Books 
Lopez y Royo, Alessandra (Iyer) 
1997 Prambanan. Sculpture and Dance in Ancient Java.  A Study in Dance 
Iconography. Bangkok: White Lotus 
1999 " Indian or Javanese? Dance vocabulary, movement interpretation and cross-
cultural interaction" Seleh Notes, the UK gamelan newsletter,  6, 3 pp10-12 
2000 “ Nawanatya: understanding dance and dramatic performance in ancient Java” 
in Lokesh Chandra (ed) Society and Culture of Southeast Asia. Continuities and 
Changes. New Delhi: International Academy of Indian Culture and Aditya 
Prakashan [Sata –Pitaka Series vol 395] pp 103-119 
2001 "Sculpture, dance and heritage: animating dance sequences from temple 
reliefs using movement modelling software" Internet Archaeology n.10, University 
of York http://intarch.ac.uk/journal/issue10/iyer_index.html 
2004 “Issues in Dance Reconstruction: Karanas as dance texts in a cross-cultural 
context” Dance Research Journal 36,2 pp. 64-79 
 Ketut Rota 
1991 Gerak Tari Bali Dilihat Dari Istilah-Istilah Yang Dipergunakan: Studi 
Eksploratif Pada Legong Kraton Lasem.  Laporan Penelitian.  STSI, Denpasar, 
Bali. Unpublished dissertation.  
Leigh Foster, Susan  
 14 
1995 Choreographing history  in Choreographing History: 3-24  ed. Susan Leigh 
Foster.  Bloomington: University of Indiana Press 
  
Naidu, V.N and Naidu, V.S and Pantulu, V.R  
1936 Tandavalaksanam or the fundamentals of ancient Hindu dancing Madras: 
n.p 
O’Shea, Janet 
2000 Technique and Theory in the Work of Kapila Vatsyayan Dance Theatre 
Journal  32, 1:  82-102 
Pollock, Sheldon 
1996 The Sanskrit Cosmopolis, 300-1300: transculturation, vernacularization and 
the question of ideology  in ed.  J.E.M. Houben  Ideology and status of Sanskrit 
Contributions to the history of the Sanskrit Language Leiden, New York and Koln: 
E.J. Brill 
 
Rubidge, Sarah 
2000 Identity and the open work in Preservation Politics. Dance 
revived,reconstructed, remade : 205-216 ed. Stephanie Jordan London: Dance 
Books 
 
Sedyawati ,Edi 
1982 The question of Indian influence on ancient Javanese dance Rima 1 6, 2 :59-
82 
 
Shanks, Michael 
1997 (with Lampeter Archaeology Workshop): Relativism, objectivity and the 
politics of the past Archaeological Discourses  2:164 –198 
 
Shanks, Michael and Tilley, Christopher 
1992 Re-constructing archaeology. Theory and Practice  London and New York: 
Routledge [1st edn 1987] 
 
Soedarsono, RM.  
1990 Wayang Wong. The State Ritual Dance Drama in the Court of Yogyakarta. 
Yogyakarta: Gadjah Mada University Press [1st edn 1984] 
 
Subrahmanyam, Padma 
 1978 Karanas in Indian sculpture. Unpublished PhD thesis, Annamalai University 
 
Vatsyayan, Kapila 
1968 Classical Indian Dance in Literature and the Arts New Delhi: Sangit Natak 
Akademi 
1983 The Square and the Circle of the Indian Arts New Delhi: Roli Books 
International 
1983a Dance sculptures in Sarangapani temple Madras: Society for 
Archaeological Historical and Epigraphic Research 
Wright, Astrid 
2002 Review of Alessandra Iyer Prambanan: sculpture and dance in ancient Java 
Asian Perspectives 41,2 : 174-178 
 15 
 
 
 
Non-print  
Balakrishnan, V. 
1992 Bharatiya Natya Sastra Documentary Series for Doordarshan, New Delhi 
[video-recording].