Lakshmi Amman's Homepage

Ancient/Medieval Indian Dance Texts

Tracing through the history of historical texts and their translations is its own research project.  When working with the publications available today, a couple of points become clear as a general way of thinking about these publications and the original sources they represent.

First, a translation is an interpretation

The translators/editors have to make their own decisions about word choices. Generally English vocabulary and grammar is quite different than any of the source languages.  Describing how a body should move or be arranged in a single language is hard enough (take a look at how dances are annotated in any language - all annotations require some form of living training or leaps of faith to interpret) - it's much harder do it across languages.  Translators can also be a mix of skill sets.  Some come from a linguistic background and have a core expertise on the wide body of Sanskrit literature, or on Sanskrit performing arts and historic dramaturgy.   Other languages/cultures deserve similar focus - but Sanskrit gets the widest body of academic research similar to how Latin does in the Western world.

Other researchers combine language and a deeply invested Indian classical dance background with a goal of contributing to the body of knowledge in their subject matter expertise.  In these cases, particular body positions and choices of words can be far easier to interpret as a dancer, however, be aware that the background of the translator can contribute a bias to the translation.  There are variations, generally minor ones, between the execution of movements and positions between dance forms, and this can lead to an assumption around what a vague turn of phrase can mean.

Translators must choose what they translate

There can be multiple, contradictory, copies of a given historical text, and contradictons within the text itself.  The translator needs to decide the scope of the translation and what to do about contradictions.  Even within the historical context, a given treatise may get commentary and later additions (ascribed to the original author) - which a modern translator must figure out how to treat.  Also, the more we study, the more we know - a given translator can release multiple editions of some of these books, potentially with serious revisions.  The efforts to begin modern translation and publication stretches back to pre Mughal eras (before 1526 A.D.), and the state of the art has been continuously moving forward as translators build upon each other's work and their own as their understanding grows.

A note on language

Where the language of the original source text is not mentioned, it is safe to assume that the language is Sanskrit.  The work of translating, analyzing and publishing information on Sanskrit texts about dance is far ahead of any other language in the Indian subcontinent, at least in English.  Newer works are starting to cover other languages - for example, the History of Tamizh's Dance by Dr. S. Raghuraman covers a copious number of texts written in Tamizh that reference and describe dance in detail.  Hopefully over time, I will be able to find more and add to this list.  A way to look at India is to think of it as analogous to Europe.  If you're thinking of it as "Spain" - you're thinking too small.  India is a mix of religions, cultures, languages, philosophies, tribes and empires.  And it always has been.  As such, there are plenty of places on the subcontinent where Sanskrit and the concepts espoused by Saskrit dance writing were influential, but not dominating.  For example, in many Dravidian language based cultures, we see the existence of Sanskrit words and concepts, but not as the Sanskrit texts dictate them.

Dates - an art of their own

When it comes to looking at the texts themselves, judging their actual dates can be even harder.  While many texts have an attributed author, it's quite common for works in India to be attributed to an author when the work is truly a disciple or other intellectual heir of the author who may even be centuries removed from the named person.  So an author can almost be considered a group of people.  A work can also be layers of publications with early chapters and later additions - so which date is most accurate?  The earliest date when any chapter is thought to be written?  or the latest likely date of the last chapter?

Very few extant written artifacts remain - the nature of India is not conducive to maintaining organic materials like palm leaf manuscripts or paper.  And the eras of conquest (Mughal and British) caused further destruction.  Most dating is achieved through linguistic analysis based on the style of the language, the word choices, and layers of references to other texts (ie if Text A refers to Text B, it seems reasonable to assume Text A was written after Text B, so if we know the date range of A, B must be earlier... or at least part of B must be earlier).  There can be layers upon layers of inference here where a single new piece of information can change everything.

... So far.

To keep an already very long page just a tiny bit shorter - I have refrained from putting "so far" at the end of almost every sentence.  That does not mean anything here is certain.  When I started researching this stuff 20 years ago, the web pointed to 1 historical treatise (the Natya Sastra).  Within 5 years, I'd found 5, and published my first mudra page.  Now I'm working on a web app powered by a relational database simply to re-publish that same material (ie, the state of the art of what we know about dance positions pre-1650).  The state of the research is increasing day by day - but even more so, the availability of texts that were translated 50-60 years ago in arcane sources, are now available in PDFs or at least in arcane Internet book searches.

So... when I say that something "is not available" or "does not exist" - please put a "yet" in there.  It's not YET been found.  Or "we have no examples".  This is the state of our knowledge as of May 2022.  

Deducing the code - Sanskrit for new Dance Researchers

If you look at this list, you'll see many of these titles are formulaic.  Many are very intuitively named, but it's not intuitive if you don't speak Sanskrit.  Here's a primer - note, all these words have a depth of meaning that transcends what is reasonable to put into a single web page:

 

What's in the rest of this page?

This page is one part overview, one part shopping list/inventory, and one part time line.  It started as a way of putting the books in my library in context, and evolved into a quick summary of dance texts over the SCA period in the Indian subcontinent.  Full disclosure - I have omitted texts that seemed to be either purely about music, or about drama in a way that was so tangential to dance that I thought it would not give me much information in terms of dance reenactment.  If you want a fuller picture (and yet still quite readable) - I cannot recommend this work highly enough - "The Evolution of Classical Indian Dance Literature: A Study of the Sanskritic Tradition" by Mandakranta Bose (Somerville College, 1989).  I am so thankful to have found it, as this research puts dance and dance/drama theory into a full context with a great deal of thought and explanation into how concepts evolve with great context.

Key

= This text is included in the Dance Treatise Project (or will be soon).

(click any box to expand for more information)

We have no extant examples of these.  It's entirely possible that some of these verses are repeated or redacted into the Natya Sastra.  Authors are mentioned as by Silalin and Krsasva.  But the way we know about these writings is from other works referring to them as explanations of dance.  In particular, the Aṣṭādhyāyī by Pāṇini, a Sanskrit grammar text. (Bose, p. 3)

The language is Tamizh - the language of the area modernly known as Tamil Nadu.  The Westernized version of this word is "Tamil".  The language is quite ancient going easily back to the 5th century but like any language, it evolves over time with old, middle and late Tamizh being distinct evolutions.  This is a Dravidian language, meaning its structure comes from a Dravidian base, rather than the Indo-European base of Sanskrit.  That said, there is enough crossover in terminology that one can see Sanskrit dance words used within the Tamizh grammar.

The period of this writing is called the Kadich-Cangam period, and is also known as the period of king Mudath Thirumaaran.  This work is referred to in the well known text the Silappadikāram (estimated date of composition is 171 CE).  There is a contrary view here - Wikipedia cites it as a 10th century text that is mentioned in a commentary of Silappadikāram written at about that time.

Panchamarabu includes the same sorts of formats that we see in the Natya Sastra - lists of particular postures, combinations of postures forming karanas, and the same kind of segmentation of the body into parts (head, eyes, body, hands, legs).  Something notable is that is also includes at least a lightweight description of dances among groups of people, dances using certain props (sticks, air bellow??, rope, ball, bow, leaves and "draping like a female" (quote from S. Raghumaraman p. 14)).  Also dance where the performer dresses up like foreigners (as in people from other countires) or gods & goddesses.

There are some internet references to an English translation of this work but as of yet, I have not been able to find it for sale.

This text has a wild history - it is mentioned in Adiyaarkku Nallar's commentary on the Silappadikaram in the 13th century.  But it's mentioned as being "lost".  In the time since it has been found.  I don't know the story.  Based on that, alone, it's safe to put the date on this as before 1300 (end of the 13th century) but I can't honestly guess on long it takes to lose a dance treastise that is still within living memory of the writer.

This is also a Tamizh text (see Panchamarabu), and it's author originated from the town of Kootanoor.  He could also be a contemporary of Tolkappiar which would make this text as early as the 5th century.

It sounds like there is enough specific reference on movements to make this an interesting read.  I have no indication yet that a translated version in English exists.

The earliest available comprehensive work on dramaturgy.  It includes dance - and much, much more - describing all aspects of theatrical production, of which dance is a significant part.  Predecessor works that inspire this are referenced in literature, but to date, have not been found and/or translated.  It is also the foremost source generally cited as the historical reference for all modern classical dance styles.  That said, it's the inspiration for all modern classical styles, but the strict cannon of none of them.

There is a very wide range of dates attributed to the Natya Sastra.  The early range of dates are based upon linguistic dating, asserting the time range as 500-300BCE because the words, figures of speech, and location names are relevant to that time period, and no longer used at later periods.  The later dates of 700-900CE are an estimate based on linguistically dating the latest chapters in the 36-37 chapters available now.  The analysis around what date it, or all of it's parts, may be deserves it's own publication.  A smaller, reasonable bet is 200BCE-200CE based on Bose, who has done a credible and large chunk of research on this topic.

The text is written in Sanskrit and attributed to the sage Bharata, but this word gets overloaded in a number of different directions.  It's also likely that many authors contributed to this work over a longer period than a single lifetime.

We also know this isn't the only work of this time - other Sanskrit authors of the period refer to other works.  A particularly tantalizing work is Kohala's, which is the most referenced and seems to be held in high esteem, but the only remnants we have are quotations from other sources (so far?).

 

This is a Sanskrit text on everything.  It's got 3 volumes, and the one relevant to dance is volume 3, which includes dance as wel as many other arts.

To the best of my knowledge, there is 1 English translation.

Possibly even more mysterious than the Bharatārṇava.  No one seems to agree on a date on this one, and the text is not easy to parse into a date.  Also, Narada is an author with no clear life, or time period, and seems to be an iconic name referenced in many treatises across a wide period.

It's also just as elusive.  A Sanskrit copy published in 1920 is available in PDF form.  Unfortunately, that's not helpful.  This is particularly tantalizing as Bose describes the dance portion as "This seems to be a desi tradition that was obviously not very highly regarded-possibly because of its vulgar eroticism-since it appears in no other text." - I would love to know more about vulgar eroticism!  It's also described as very hard to read and it sounds as if the dance information here is minimal compared to the music and general performance information.

Another Tamizh text.  This one does not have a known author and the dating on it vague but based on linguistics and style.

This may be more of a theory/philosophy book on drama and dance, rather than a book with specific listings of dance positions.  But the way it lays out thinking on dance is unusual and worth studying. 

I have not yet found an English copy (or any copy) available for sale or download.

This is really interesting as a text for the concepts it introduces, according to Bose.  However, it's also described as a text that is primarily on dramaturgy, and there's no mention of specific dance movements being in here.  But don't take my word for it - there's a copy here.

The Natya Sastra is now old enough and respected enough that sages are growing the knowledge of the age by commenting on it.  These can be really interesting, as the contrast or further color coming from a verse by verse commentary can give both how dance is thought of in the era of the commentator and also how the older text is now being interpreted and responded to.

This is considered an important work, although it's state is noted to be corrupted and hard to trace through.  What is great about it, is that gives a sense of how dance changed between Bharata's time and Abhinavagupta's.  Abhinavagupta also quotes other authorities extensively, giving a wider range of the thought of the period.  Abhinavagupta was from Kashmir.

Another work on dramaturgy.  This isn't interesting for the specifics of dance, but for the definition of Sringara and the evolution of this as a major concept in dramaturgy.  This is a huge thing in modern Indian performance, and also a tricky aspect of the Indian Independance movement and how Indian dance is performed and interpreted today.  This is the time period where thoughts on this topic are beginning.

The author is presumed to be from modern day Kerala.  This text is is associated with the styles of Kerala, including Kuchipudi and Katakali.

I don't have much else on the history of it, but I did find an online translation.

 

Another case with no English translation that I have found.  But this does include annotations of movements, as well as new discussions in drama aethetics.

This one of the trickier texts to date.  Ghosh argues for 600-1247, while Bose makes a case for 12th/13th century.  Also, the assertion of Nandīkeśvara as the author.  There's no clear date on when this sage lived, and he's given credit for two texts - this and the Bharatārṇava - which disagree with each other.  Also his writing seems to span more than a human lifespan.  When dd he live?  Is he real?  the jury is still out.

All that aside - this is a fanstatic source.  I've worked with several different copies of it, and they are generally readable and comprehensible, but the text is short enough to be approachable.  It's fair to say that this is stronger source material, in many ways, the the Natya Shastra for Bharata Natyam, at least.

The writings of Mānasôllāsa are a great text for everything.  It's literally an encyclopedia of its day.  It's written in the Western Chalukyan dynasty which is in modern day Karnataka and Andhra Pradesh.  This a truly broad and yet detailed text covering a number of elements of life at this time.  One volume of this set of tomes is the Nṛtyavinōda.  It returns to the same type of format of the Natya Sastra in that lists sets of movements of each part of the body with descriptions of how to execute them and what they mean.

This is the first text that focuses on dance and music, and not as adjuncts to drama.  But what truly caught my attention on this text was two things - 

  1. Bose describes it at having a good set of descriptions of desi dances, which are coming into their own in this period
  2. It couples dance and the talas for dance both margam and desi - which would be a great new find, for me at least.

This is a case where the scholars are trying to date the life of the author by the state of the text.

As can be seen from this timeline, there's a great burst of extant texts at this time period - and there is a trend of both referencing that which came before (back to the NatyaSastra) which helps to put the works into a context.  This is not a work I've been able to find, but based on Bose's research, it sounds as if at least one chapter is devoted to dance, although it sounds like it's devoted to an auxiliary role.

The most interesting part, to me, is a tantalizing reference that the teaching of dance might be discussed.

I haven't gotten a conclusive date on this.  Only alluring suggestions that this text covers dance composition (that would be a first for my research) and that it's concepts seem later than the Abhinaya Darpana.  These two texts are attributed to the same author, although that seems unlikely.

I have yet to find a single case of an English translation to this one.  It's in print in Sanskrit, and referenced in great detail in several texts, but I have yet to get my hands on a reasonable copy.

The empire of this text is the Yādava dynasty.

This is one of those texts that EVERYONE refers to.  Often a text on music (or occasionally dance) will reference back to this.  It is described as both methodical and clear, so no surprise.

6 out of 7 of the chapters of this text are about music.  The 7th, which is about dance, seems to draw heavily from the Abhinayadarpanam and other sources.  Unfortunately, the copy I have, gives the first 4 chapters, and plans to publish a third volume with chapters 5 & 6, with no mention of 7.  This is killing me a bit because Bose writes (p55):

The seventh chapter, on dance, is the one that is of interest to us. It has 1678 verses and is an excellent source of our knowledge of both marga and des~I traditions of dancing in the author's time. In describing the marga tradition he follows Bharata but he extends our knowledge significantly by giving for the first time a systematic account of the desi tradition of dancing. Though the Manasollasa and the Sangitasamayasara had previously introduced the desi style into the traditional account of dancing, it is Sarrigadeva who systematizes that account.

From what I can tell, no English translation of Chapter 7 exists as of yet.

The Nṛttaratnāvalī by Jāyana (also called Jāya Senāpati) is in Sanskrit, and written down in Telegu text.  It dates from the Kātatīya Dynasty, which covered what is now modern day Telegana.  The rulers were worshippers of the goddess Kātatī and this was a golden age for the region, with a lot of writing on the arts and other intellectual concepts.

Jāyana wrote 3 texts on dance, however the Nṛttaratnāvalī is the one that survives.  The Gītaratnāvalī and Vādyaratnāvalī did not survive.  The work divides itself into 8 chapters - 4 devoted to marga - the classical format that traces to the long linneage of historical treatises, and 4 desi dance which isn't within this cannon and is generally described as regional and popular, which the author then describes meticulously.  This kind of division starts at this time period.

It's worth noting - the authors of the Saṇgītaranākara and Nṛttaratnāvali are contemporaries and yet make no mention of each other.  One focuses on music, and one on dance.  But they are from neighboring states, and it may very well be a rivalry.

Another case of ?who came first? Sangitacandra or this?  We have to admit, that many works of a same century are simply contemporaneous.  There's enough specificity on movement (although possibly a repetition of other sources) to make this interesting.  But the standout appears to be the description of staging battle scenes.  I admit to being quite curious on what medieval Indian fight choreography would look like.

Earliest text to describe the desi dance movement named kalasa.  There are references earlier, but this includes sets of movemetns and seems to have it further developed. 

Written by a Jain writer (not the first in this list, BTW), and an abridgement of a longer work that has no extant copies, this is what remains.  Unfortunately, according to Bose, there's nothing new here.  It's a good, organized summary.

But also worth noting that the thoughts are alive and well, and largely unchanged, 50 years later or at this time - depending on how you intrepret these ranges of dates.

A text which is purely devoted to hand (hasta) movements as the name suggests.  The author describes it as "very elaborate", and I would have to agree.  The number of verses per gesture seems to be among the highest that I've seen.  The author also mentions the influence of the Natya Shastra and Saṇgītaranākara, but very little influence from the Abhinaya Darpana.

The best I can tell, the work is affiliated with the "eastern tradition", although I know very little of what that constitutes.  I also see the author most often referenced as part of the Bengali literary tradition.

The earliest date is based on dating texts that are referenced by the author either in this work or in his later work (the Saṅgīta-dāmodara), which at least places the author's life span.  The latest date is based upon the approximate dates of other texts, which reference this work, or the Saṅgīta-dāmodara.  Among the earliest of these is a work written by Sukladhavaja, who is known to have died in 1571 AD.

Interesting in the mention that they cover revolving motions and leaps.  Two things I have not seen elsewhere.

The author is believed to be from east Bengal.  It sounds like a fairly comprehensive treastise on dance, music and dramaturgy that gets often requoted. 

The same author published Hastamuktavali - which deals only with hand gestures.

I have yet to find an English translation to either.

According to the author, this was written to please the Emperor Akbar, although the author is believed to be from Southern India.

This is 4 chapters, with a missing 5th, and the 4th is devoted to dance. It's quite comprehensive, includes both desi and margam and Bose suggests it connects ancient dance to the modern era.

Śrīkaṇṭha is a contemporary of Puṇḍarīka Viṭṭhala, and in fact puts himself as Puṇḍarīka Viṭṭhala's student.  He was a court poet of Satrusalya (Jam Sattarsal of Navanagar near Dvaraka).  It covers a number of different regions of dance.  This text also is said to mention the pushpanjali or flower offering, and I am terribly curious to see how it compares to it's modern presentation as this is still an important part of classical dance.  It also mentions details rules for makeup which Bose mentions are new with this text.

Unfortunately, I have not been able to find an English translation.

The author was in the court of Jahangir.  It's got 1 out of 7 chapters on dancing, the last.  It includes movements and stylistic information but does borrow heavily from the Nartananirnaya.

I have not yet found an English translation.

The text is only partial, but the description from Bose is alluring in that it talks about sequences and possibly dances.  It's esoteric enough that it's even hard to find references to it in Google.  Notice - there are two Saṇgītamakarandas - this is not to be confused with the 700AD-900AD version by a different author.

There is new material in this.  It's hard to tell from Bose how interesting this would be for reenactment.  Also, I have yet to find specific dates, which would give some idea how easy it would be to put this work within the SCA period.

This is not really included as anything but an aside in Bose's study.  This text is strongly connected to Odissi dance and is written in Oriya.  Dating it is quite rough.  The two translations I have put the most plausible date in the range of 1670 (the date the author says he published it) to 1770 (the time period of when the author's patron was alive and in power).  The inconsistency on the author's part could well be because there were mutiple authors, so my bet is that it's approximately this range.  

Others do put the text at 12th century or 15th century - likely based on linguistic analysis, but I haven't seen particular details that help me understand why these are the more plausible dates.  In his translation of the Śrīhastamutāvalī, Maheswar Neog attributes this text to the 12th century with no other details.

After these texts, we start to get beyond the target time frame of the SCA, but the writing continues.  The scope of this list is long enough, but other post 1650 treatises include: