Chapter 4
In Search of an Ancient Kathak1
Many histories of Indian dance make claims of Vedic origins. These histories begin in ancient times, usually combining the mythological story of the creation of dance for the Gods’ amusement with archaeological evidence like the tiny ‘dancing girl’ statue from the 4,500-year-old civilization at Mohenjo Daro (see Bhavani 1965, Vatsyayan 1974, and Khokar 1984 among many others). Moving quickly through the millennia, the next stop is the seminal treatise Nā
yaśāstra, which dates from the second or third century CE. The aesthetic and choreographic material in the Nāyaśāstra is often seen as the basis of all Indian classical dance, with some authors postulating that ‘originally there was … only one art of dance in India and that, as time went by, this matrix splintered, lending itself to regional adaptations’ (Khokar 1984: 53; also see Vatsyayan 1974: 2). While one can easily dismiss this need to link present-day culture with antiquity as part of the Orientalist and nationalist movements discussed earlier, its ongoing presence not only in the literature but also in dance training and publicity material makes an examination of such claims a crucial beginning to any study of dance history in India.
There are certain issues that arise specifically when attempting to place kathak in the ancient past. Although one can find dance websites saying that kathak is 4,000 years old, there is little if any evidence at all that supports this. Dance scholars like Narayan (1998) and Srivastava (2008) have attempted to connect kathak to ancient statues and treatises, but few of the movements and gestures in kathak have anything in common with the choreographic details preserved in ancient texts or sculptures. Researchers have therefore largely found themselves with the tasks of considering the contexts, imagining or attempting to recreate the movement patterns, and then deciding if anything described in writing or sculpture seems to relate to kathak in any way. As discussed in Chapter 2, however, the name of the dance denotes a form of storytelling and offers a link to the Sanskrit term kathakā or storyteller. Referring to appearances of the word kathakā in Sanskrit literature has therefore formed the basis of the claims of kathak’s antiquity, rather than attempts to connect the dance to ancient dance postures.
Mahābhārata
The most widely disseminated example of this are the ancient Kathakas mentioned in the Mahābhārata, ‘who recited Kathās or stories from the sacred Puranas and epics with expository gestures and dance’ (Devi 1972: 166). It is important to look at this claim in some detail, as the belief that there are ‘Kathavacaks in the Mahābhārata’ has entered common knowledge and can be brought up, sometimes quite aggressively, in both conversation and scholarly discourse without any need for verification. The Mahābhārata, of course, is not a dance or music treatise, but an epic tale of warring families and divine intervention. Like other such epics, the Mahābhārata existed orally long before it was compiled and written down in Sanskrit sometime around 300 CE. The story of the Mahābhārata is a common subject in many forms of Hindu theatre; key parts of the epic are still re-enacted in dance-dramas, puppet-shows and both vernacular and professional theatre. In kathak today, the story of the dice game and the intervention of Krishna in the assault of Draupadi is the basis of one of the most commonly performed gat bhāvs.
As well as providing dramatic story lines, it is claimed that the Mahābhārata contains two apparent references to kathak. In the first book, the Ādi parva, section 206, line three contains the following text: ‘kathakāś cāpare rājañ śrama
āś ca vanaukasa
/ divyākhyānāni ye cāpi pahanti madhura
dvijā’ (Mahābhārata 1.206.3). This passage, which lists the companions of the hero Arjun on his exile into the forest, was translated in the late 1800s by Kisari Mohan Ganguli as: ‘narrators of sacred stories, devotees leading celibate lives, Vanaprasthas, Brahmanas sweetly reciting celestial histories’ (Ganguli 1883–1896).2 Norvin Hein, in Miracle Plays of Mathura, interpreted this passage thus: ‘kathakas and forest-dwelling ascetics, and brahmans who recite sweetly the divine tales’ (Hein 1972: 51). In Rhythmic Echoes and Reflections, Narayan translated it similarly: ‘Arjun, on his departure for the forest, was accompanied by an entourage of Kathaks and forest dwelling ascetics and brahmins who recited sweetly the divine tales’ (Narayan 1998: 8–9). The thirteenth book of the Mahābhārata, the Anuśāsana parva, contains another tiny section that is referred to: ‘gāyanā nartakāś caiva plavakā vādakās tathā / kathakā yodhakāś caiva rājan nārhanti ketanam’ (Mahābhārata 13.24.16). This fragment is part of a much longer list of Brahmans who, through various sins and misbehaviours ranging from contracting serious diseases to accepting regular pay for worshipping, are not to be invited to attend sacred ceremonies such as the śrādha or funerary rites. The excerpt in question concerns performing artists and Ganguli translates it as: ‘Those Brahmanas that are, by profession, vocalists, or dancers or players or instrumental musicians, or reciters of sacred books, or warriors and athletes, do not, O king, deserve to be invited’ (Ganguli 1883–1896).3 In the dance literature, however, the word kathakā is again left untranslated: ‘Singers, dancers, rope dancers, instrumentalists, Kathaks and fighters are not to be invited, O King’ (Narayan 1998: 9, cf. Hein 1972: 51.).
Narayan and Hein were the only authors to provide the actual citations from the epic, but as mentioned above, references to the ‘ancient Kathakas’ who travelled the countryside in Vedic times reciting sacred stories and epics are standard fare in almost all the books on kathak dating after 1965 (see for example Bhavani 1965, Singha and Massey 1967, Devi 1972 and Kothari 1989). Earlier writers do not make such claims. La Meri (1964 [1941]: 17), for example, described Kathaks who used to recite epic poems like the Mahābhārata, but did not give any indication of when in history this took place, nor did she make the claim that Kathaks actually could be found in the epic. Vyas, in his 1959 article on ‘The Background of Kathak’ for Marg magazine, included what is perhaps the first reference to any Vedic connection, but his language seems purposely vague (1963: 5). Like other aspects of kathak historiography, the person who originally located these tiny citations remains a mystery and most of the earlier twentieth-century sources make no connection to Vedic times at all. Yet even though the appearance of the references in the literature is difficult to trace, the citations themselves can be examined. Two questions arise concerning translation, and both lead to further queries concerning interpretation.
The work ‘kathak’ in Devanagri script has two spellings: 
(kathak) and 
(katthak). The first spelling is by far the more common but the second version (with its doubled middle consonant) is commonly considered older. Some books use the second spelling (for example Kothari 1989 and Kapur 1997), but perhaps only with an eye to seeming more rooted in ancient literature. In Sanskrit, however, the two spellings may indicate different root words with somewhat different meanings.4 The root word of 
is 
(kath): ‘to tell, relate, narrate, report, explain, describe’ (Monier-Williams 1993 [1899]: 247). In Hindi, the word is similarly translated ‘to tell, to recite, to expound’ and also ‘to compose (oral poetry)’ (MacGregor 1993: 162). This root gives rise to a host of derivatives – kathan: a narrative, kathit: said or uttered, kathā: story, tale or legend – and is prominent in the names of other performing arts such as kathakali, the dance-drama of Kerala, kathputlī puppetry from Rajasthan, and of course kathāvacan. This link of the term kathak, both in reference to a dance and to a caste of performers, to storytelling is a crucial part of kathak’s identity. The second spelling, however, indicates the root 
(katth), which in Sanskrit means to boast, to praise or celebrate, to flatter or coax, and even to abuse and revile (Monier-Williams 1993 [1899]: 246). This word does not appear in Hindi or Urdu Dictionaries although the third translation of kath in the Oxford Hindi-English Dictionary is ‘to reproach’, which may point to this second root and meaning (MacGregor 1993: 162). More to the point is the observation that although some books use the word 
(katthak), none makes any mention of the different meaning and interpretation brought to mind by this spelling. The term evokes, not a community of storytellers recounting sacred epics, but a group whose profession is more reminiscent of bards, satirists or praise-singers.
The other question of translation that emerges is more straightforward. Assuming for the sake of argument that the 
(kathak) spelling is correct or that the difference does not matter, one finds the following definitions and translations:
kathak, and H. katthak, and kathik, s.m. Narrator, relater, reciter; one who publicly reads and expounds the Purāas; – a professional story-teller; a kind of singer or bard; chief actor; a dancing boy (Platts, Dictionary of Urdu, Classical Hindi and English, 1997 [1884]: 813).
kathakā m. A narrator, a relater. 1. A chief actor, speaker of a prologue. 2. A disputant. 3. A story-teller (Apte, The Practical Sanskrit-English Dictionary, 1957 [1889], vol. 1: 525).
kathak [S.], m. 1. narrator; specif. public narrator and expounder of sacred legends. 2. drama. The speaker of a prologue, an actor. 3. a community of singers and dancers. 4. name of a style of dance (MacGregor, Oxford Hindi-English Dictionary, 1993: 163).
The emphasis in the two nineteenth-century dictionaries is clearly on a ‘kathak’ as a public speaker, whether a narrator, storyteller or actor, and the identification of kathak as the name of a community or a dance style is only in the much more current Oxford Hindi-English Dictionary. Translating kathak, katthak or kathakā simply as ‘story-teller’ or ‘reciter’ as Ganguli does without any implication of caste or community gives rise to a more general and arguably more accurate interpretation of the lines in the Mahābhārata. Arjun was accompanied into the forest by storytellers or bards. The ‘forest-dwelling ascetics’ are not associated with any particular present-day group. It therefore seems very likely that the storytellers are also being identified simply by their activities rather than by the name of their community. The other citation supports this further. In the list of unwelcome entertainers, Kathaks are the only group identified by name in the translations. Translating the passage as ‘Singers, dancers, rope dancers, instrumentalists, storytellers, and fighters are not to be invited, O King’ actually makes more sense. To interpret the word kathakā as denoting a distinct community of professional storytellers – one that can be connected to the birādarī of today – begs the question of whether the singers, dancers and rope dancers ought to receive similar scholarly attention. Yet, the context of the passage indicates that these are occupations rather than the names of communities and it is worth noting that scholars and translators like Ganguli who have no connection to the kathak world and its need for ancient roots do not see any correlation between the passages and present-day performing artists.
The scattered later citations identified by kathak scholars also offer little if any useful information in uncovering the history of the kathak dance and cry out for a less preferential approach to translation. Hein (1972: 51–2), Kothari (1989: 1–2) and Narayan (1998: 9–10) each drew attention to small passages in writings from the tenth to thirteenth centuries that mention Kathakas, but it is still difficult to see these as offering any sort of convincing connection to today’s dancers. To argue that the ‘kathakā Sangataka’, for example, who entertained King Sahasranika with a long secular love story one evening in Kathāsaritsāgara (X.2, cited in Hein 1972: 51) was a member of the Kathak community seems far-fetched and certainly the descriptor kathakā is just as easily translated as narrator. At any rate, identifying a Kathak recounting a love story in an eleventh-century Hindu court undermines the version of history that places Kathaks in temples and attributes the secularization of the art form to the Mughals. Srivastava has taken the search even further, finding kathaka and other cognates in a number of sources including Jain lexicons, the Ganapātha, the Ganeśa Purāa, Amarakośai and the Śabdakalpadruma (2008: 23–6). Yet, one can only reliably conclude from her search that there were people in pre-Mughal India who related narratives, and the Sanskrit word ‘kathakā’ was used to refer to them.
Scholarly evidence linking kathak or the Kathaks to Vedic India or even just India before the thirteenth century is therefore spurious at best, and is arguably derived from twentieth-century politics and the search for ancient origins rather than any historical facts. Certainly, there are enough inconsistencies in the various meanings and two spellings to call the accuracy of the connection into question. The contemporary word ‘kathak’ (or Kathak) is obviously a cognate of the Sanskrit word denoting someone who recited or narrated stories, although they were not necessarily sacred (as the reference in Kathāsaritsāgara indicates). The claim that appearances of the word kathakā are evidence of a connection between today’s dancers and the narrators of the past, however, must be seen as either highly imaginative or so broadly true as to be meaningless. There were dancers, singers, acrobats, fighters and storytellers in the past as there are today, but the older Sanskrit material seems to refer to them by the name of their activity rather than by the name of their community as later material does. The contexts and the sparseness of the references support this assertion. If there was a caste or a community of people called Kathakas that played a significant role in ancient Indian society, surely it would be somewhat more visible in the literature. Yet rather than visible, the Kathakas become invisible as even these scant references become rare in documents specifically concerning music and dance.
Sanskrit Treatises on Music and Dance
Extant music and dance treatises in Sanskrit range from the oft-cited Nāyaśāstra, believed to have been composed sometime in the second or third centuries CE to the seventeenth-century Nartananirāya written during the reign of the Mughal Emperor Shah Jahan. The Indian dance world places the greatest importance on the Nāyaśāstra and another early work, the Abhinayadarpaa, which dates from around 1000 CE. Recognized as the earliest extant treatise containing material specifically describing dance, Nāyaśāstra includes not only much of the terminology and aesthetic philosophy of dance now used by dancers and dance scholars but also long lists of codified postures, gestures and movements that were then reiterated and expanded in subsequent treatises. Abhinayadarpaa focuses, as the title suggests, on abhinaya or expressive gesture, and comprises descriptions of glances, movements of the head, eyebrows and neck, and hundreds of hand positions culled, according to the text, from other books on gesture. Students of Indian dance are often taught to recite the names of the hand gestures or mudrās while demonstrating them, and study of both texts usually forms an important component of advanced degrees in dance and choreographic training.
The role of this ‘ancient’ movement vocabulary in today’s dances, especially in South Indian genres like bharatanāyam, needs to be seen in the context of the Indian classical dance revival when material and terminology from these treatises were imported into contemporary practice (Meduri 2005, O’Shea 2007: 35–7; see also Chapters 8 and 9). Publication and dissemination of various treatises, first in Sanskrit and subsequently in translation, became more widespread through the 1890s and early twentieth century. As Indian culture was reclaimed and revived between the 1920s and 1950s, the connection of music and dance to Sanskrit descriptions contributed to a process of classicization. Early publications such as Devi (1920) that reveal the long history of dance in India and aim to connect past with present have been critically examined to some extent through more recent work. Although common knowledge among dancers still unquestioningly attributes the origins of Indian classical dance to the Nāyaśāstra, Kapila Vatsyayan had already pointed out by the mid-twentieth century that ‘the history of contemporary classical dance styles in their present form cannot be traced too far back, … [although] the links between these dance styles and the earlier tradition exemplified in literature and sculpture unmistakably exist’ (Vatsyayan 1968: 364). Recently, Mandakranta Bose has been even more adamant, stating: ‘the common impression that contemporary classical dances have evolved directly from the Nāyaśāstra turns out to be erroneous when we compare their technical details with the descriptions given in the text’ (Bose 2001: vii).
Most importantly for this study, neither the Nāyaśāstra nor the Abhinayadarpaa include names of dances or communities of dancers and thus the efforts to connect contemporary dance to these treatises has focused on the postures, gestures and movement patterns. Yet, this material has little in common with the choreographic vocabulary of kathak. Some of the mudrās, or hand gestures, have found their way into kathak but there is generally a recognition that although gestures such as patākā (formed with a flat forward-facing palm) or muśi (closed fist) are commonly used in expressive dance, identifying them with Sanskrit names is a recent addition. Similarly, teachers will occasionally use vocabulary such as karanā (posture) or rasa (aesthetic emotion) drawn from the treatises to describe characteristic stances or expressions in kathak that have no equivalents in the Sanskrit literature.
This is generally the place of both the Nāyaśāstra and Abhinayadarpaa in today’s kathak. Gestures and other philosophical or aesthetic terminology that seem natural or can be made to fit are adopted, but in a theoretical rather than a practical sense. Although educational and documentary videos such as The Art of Kathak (Gupta 2000) and Kathak Parichaya (Maharaj 1990) contain demonstrations of the mudrās from Abhinayadarpaa and the rasas from the Nāyaśāstra respectively, it is still relatively rare that the terminology from the treatises is used with any regularity in teaching or choreography. This is summed up well by Mekhala Natavar in her study of hereditary Rajasthani Kathaks where she stated that ‘the information [the Nāyaśāstra] provides is of little relevance or concern to Kathak practitioners’ (Natavar 1997: 46 fn1) and that the ‘presumed linkage is not only outdated and unsubstantiated, but is also no longer necessary [as kathak] enjoys a status as an exemplary contemporary dance form regardless of its past’ (Natavar 1997: 5). My own observations at dance schools in India between 2002 and 2006 support this. Material from these treatises was certainly a respected part of scholarly knowledge of dance, but although it formed an important part of theory classes and exams and could be studied extensively in post-graduate programmes such as those at the Natya Institute in Bangalore, it did not seem to have any consistent practical importance in the dissemination or performance of kathak dance itself.
The Nāyaśāstra, nevertheless, influenced the treatises that followed both in structure and content. Although the Nāyaśāstra explicitly states that it describes only a central performance tradition, the text recognizes the existence of regional traditions and refers the reader to the work of another author, Kohala, which is unfortunately lost. Subsequent Sanskrit works, including the Vi
ūdharmottara Purāa (ca. 400–600) and the Abhinavabhārtī (ca. 1000), by and large follow the Nāyaśāstra in form and organization and reproduce much of the same descriptions of the central, or marga tradition (Bose 1991: 16–20 and 1995: xxiii). Matanga’s B
haddeśī, on the other hand, does include the regional or deśī traditions, and is the first treatise to use the term rāga and expand on its treatment. The Bhaddeśī is variously dated between the fourth and the ninth centuries CE, but is not complete. If it did include material on dance, it is sadly missing (for further information see Widdess 1995: 125–6 and Sarmadee 2003: 152–3). It is only in the slightly later treatises like the thirteenth century Ntyaratnāvali, and concurrent treatises on music as well as dance such as Sa
gītaratnākara (ca. 1240) and Sangītasamayasāra (ca. 1250), that one begins to find consistent documentation of the regional dance forms, although most still include the marga material taken from the Nāyaśāstra.
Works from this period deal with dance as a discrete art form (rather than simply an adjunct to dramaturgy as in the Nāyaśāstra), distinguish clearly between marga and deśī (‘classical’ and ‘regional’, as they are usually understood), introduce the term ntya (expressive dance), and begin to record dances that differ from the classical tradition described in the Nāyaśāstra. Mandakranta Bose has argued persuasively that it is in these provincial or deśī traditions that some of the roots of kathak can be found, and her discoveries are worth careful examination (for more information and clarification see Bose 1991: 5–108). One also finds the elusive Kathakas in the thirteenth-century Sagītaratnākara, where in the Seating of the Assembly, ‘story-tellers, bards, scholars, flatterers’ are to sit to the left of the king (Sagītaratnākara, VII, 1340–1350; cited and translated in Sarmadee 2003: 142; see also Kothari 1989: 1, Narayan 1998: 10, Srivastava 2008: 25). The dancers, however, are not called Kathakas but identified by various names including Acarya, Nata and Nartaka. These groups and their respective performance practices are discussed in a different section of the treatise (Sagītaratnākara, VII, 1260–1324; cited and translated in Sarmadee 2003, 139–40).5 One of the dances described in this section is called perai and includes footwork patterns called gharghara performed with ankle bells, and another dance practice contains expressive gestures illustrating a poem or kavitā (see Bose 1994: 244 and Srivastava 1994: 349–50). Other contemporary treatises such as Sangītasamayasāra (ca. 1250) and Ntyaratnāvalī (ca. 1240) describe whirling movements and spins called bhramarī and cakrabhramarī (Bose 1991: 62).
These choreographic connections are much more convincing than any links between kathak and the Nāyaśāstra and certainly more scholarly than the imaginative association of kathak or Kathaks with the Mahābhārata. Various types of spins and footwork have been central characteristics of the kathak movement vocabulary over the last hundred years, and the rendering of the poetic form kavitā with gestures and movement is still a standard performance item. None of the dance items are called kathak or connected to a dance or performers of that name, but this only calls the supposed discovery of Kathaks in the Mahābhārata into further question, rather than undermining the suggestion that these choreographic examples in these treatises may be ancestors or prototypes of material in today’s kathak dance. Nevertheless, the prevalence of rhythmic footwork and turns in many other Indian performance forms makes any claim of unbroken descent unlikely. These dance movements are much closer to kathak than any in the Nāyaśāstra, but we are still most likely dealing primarily with an inherited vocabulary rather than any dance or dances that can be called the direct ancestors of kathak.
Yet more convincing material can be found in the sixteenth-century Nartananirāya, which Bose identifies as ‘an early textual source for kathak’ (Bose 1998). Historically and contextually this assertion is on very solid ground although it must be emphasized once again that these are choreographic roots rather a single ancestral genre. The author, Pandarika Vitthala, was a scholar from Karnataka who travelled to the court of Akbar, the first ‘great’ Mughal emperor, where he wrote several treatises including Nartananirāya in the late fifteenth and early sixteenth centuries (Lochan 2002). He made it clear in this treatise that he was describing not only the marga dance of the Nāyaśāstra but also the deśī or regional dances including the dance of the ‘foreigners’, that is to say, the Persians. There are a number of descriptions in Nartananirāya that contain relevant information in the search for kathak and some particularly worth calling attention to. First among these is a much more detailed description than the account in Sagītaratnākara of the gharghara footwork. Six patterns are described ranging from apa
ava, which is ‘striking the ground with the sole of only one foot’, to the complex khuluhula, which consists of ‘striking the ground with the heel of the left foot [with] the forepart [resting] on the ground, or the rotation to the left and the right [of?] the other foot’ (Nartananirāya 603–12; translated by Sathyanarayana 1998, vol. 3: 152–5, also see Bose 1998, 55–6). The more simple steps are too general to be useful, but khuluhula can be reconstructed as a pattern similar to the footwork now called tharaghine or tharakia. In Sagītaratnākara, however, the gharghara patterns are associated with the dance called perai, which was performed by men and possibly formed some sort of comic relief (Bose 1991, 232–3 and Sathyanarayana 1998, vol. 3: 329). Thus, although there seem undeniable choreographic connections to kathak dance in the gharghara patterns (and both Sathyanarayana and Bose draw attention to this), there is no connection with the people identified as Kathaks or Kathakas, and the movement is used in an entertaining rather than a devotional context.
Two further dance descriptions in Nartananirāya may also be seen as possible roots of today’s dance. The first is a list of gatis, often translated as ‘gaits’, which are movement patterns performed by dancers that imitate the motion of animals. These include the mayūrī or peacock gati, the mainavi or fish gati, and hariī or deer gati (Nartananirāya 125–6; translated in Sathyanarayana 1998, vol. 3). The second is a dance new to India and seemingly performed by women. This Persian dance, called jakkaī, was ‘devoid of effort and action’ and performed to a song ‘sung by experts from Persia’ or even the dancer herself. During jakkaī, the dancer danced ‘with soft movements adorned by bhramarīs’ and holding her veil (Bose 1998: 52–4). Together, the gatis and jakkaī resemble portions of the kathak dance item called gat nikās, in which the dancer takes two graceful turns, now called palās, then moves in a gliding walk called a cāl that evokes a type of person, deity, mood or animal. In the ghūgha ki gat, the dancer performs a graceful pantomime imitating movements with a veil during the cāl and Bose saw jakkaī as having a clear connection to this later kathak item (Bose 1998: 56–7). Historically, the presence of these disparate bits of ‘kathak’ in the court of Akbar seem to offer some support for the claim that indigenous genres migrated to the court and were affected by the new context. None of these choreographic items is connected in any convincing way with storytellers or devotional origins of course, but if one discards the erroneous notion that today’s dance had its origins in the mysterious activities of the nebulous Kathakas, there is no reason not to see fragments of what became kathak in this evidence.
The treatises dating after Nartananirāya offer few new observations. Vitthala seems to have recorded more contemporary dances in more detail than either his predecessors or successors perhaps because, as a native of Karnataka, he was a stranger in the North Indian court of Akbar and thus interested in what other chroniclers might take for granted. Later works in Sanskrit, although containing small additions, provide little or no material relevant to the search for kathak. There is certainly no dance called kathak and after Sagītaratnākara the Kathakas do not seem to be mentioned, even in works like the Sagītacanda which include lists of types of performers (see Bose 1991: 70). Nevertheless, the details in Nartananirāya are intriguing and the document’s description of the Mughal courts and Persian dances brings us to a different set of sources.
Indo-Persian Observations
Although the birthplace of what are now known as Hindustani music and dance is usually placed in the court of Akbar, where Persian and Indian musicians, singers and dancers are said to have observed and perhaps eventually participated in each other’s performance (Wade 1998; see also Brown 2006 and Meer 2006), the sub-continent had in fact experienced prior invasions and periods of Muslim rule since the late 1100s. The supposed migration of the Kathaka tradition to ‘the Muslim court’ is therefore historically vague although it generally seems to refer to the courts of the ‘Great Mughals’, who established their presence through Babur’s conquest of north-western India in 1526. Babur’s grandson, Akbar (r. 1556–1605) solidified alliances and continued the process of Hindu-Muslim acculturation not only by marrying the daughters of Rajput princes but also establishing a court that was filled with vibrant cross-cultural exchanges. The cultural activity of the imperial court during the reigns of Akbar, his son Jahangir and grandson Shah Jahan was reproduced in the smaller courts of the Mughal nobles and eventually Rajput princes, with each courtier emulating imperial grandeur as his resources allowed (Richards 1993: 61). Darbārs (daily audiences), patronage of artists and craftsmen, and lavish feasts with poetry, music and dance as entertainment for the male guests were part of aristocratic life during sixteenth- and seventeenth-century North India (Richards 1993 and Wade 1998).
Persianate writings on music and dance from this period offer data as interesting and informative as the material in the Sanskrit treatises, but through a slightly different lens. The earliest extant treatises in Persian are for the most part translations of the Sanskrit works on music and dance discussed above and were created with the goal of explaining Indian culture to the Persianate aristocracy. Accompanying the translations are commentaries, interpretations and brief biographies of performers. Many of the authors remarked on the current performance practice that they observed and discussed how it differed from the material in the treatises, much of which by then had become obsolete. As interested in practice as they were in theory, the Indo-Persian observers left a record that potentially contains many of the details needed to fill the gap between the dances of the Sanskrit treatises and those of the nineteenth century. Yet, unlike the Sanskrit sources, which have been combed and re-combed for references to kathak or Kathakas, the extant Indo-Persian material still remains surprisingly unexamined. This is obviously in large part due to both the aforementioned Orientalist influence of William Jones and his intellectual descendants and the nationalist perspective that categorized the Mughal period as a ‘dark age’ (Chatterjee 1993: 98 and Chapter 2). Indian music and dance scholars including Delvoye (1994), Trevedi (2000 and 2012), Brown (2003, 2006 and 2010), Ahmad (2006) and Mohammadi (2006) have thus only begun to examine this rich array of sources in recent decades, and work specifically on dance is still rare. The addition of this long-neglected material is crucial to work on the history of North Indian performing arts including dance. Kathak, we are told, was shaped in the courts of the Mughal rulers and the evidence from Vitthala certainly points to Persianate as well as indigenous roots. A systematic study of dance and comparison with the roughly contemporary Sanskrit material seems yet to be done, however, as among the scholars listed above, only Madhu Trivedi has given due attention to dance. The following survey is thus very incomplete, but even the introduction of these sources provides a few overlooked pieces of the past.
The oldest of the extant Indo-Persian works, Ghunyat al-Munya (ca. 1375), is a translation into Persian of seven Sanskrit works including the Nāyaśāstra and Sagītaratnākara. In addition to providing translations of the Sanskrit material explaining music, dance and drama, the author of Ghunyat al-Munya added and inserted contemporary details, showing ‘a healthy bias for living art’ (Sarmadee 2003: xxv). Ghunyat al-Munya post-dates Sagītaratnākara by only 125 years and a detailed comparison of the two would undoubtedly be most illuminating. A quick search through the dance terms culled from the Nāyaśāstra and Sagītaratnākara reveals a combination of close translations, slight variances and occasionally widely divergent definitions. In the list of Payi (Pada) or foot movements, for example, Añcita or Anchit is explained in the Nāyaśāstra as a posture in which ‘the heel is kept on the ground and the front part of the foot is lifted’, yet is identified in Ghunyat al-Munya as a movement where ‘the heel of one foot is kept on the ground while the toe and fingers [small toes] are raised during the wheelings (Bhramarī movements)’ (Sarmadee 2003: 91). There is, furthermore, a list of seven Bhamari (Bhramarī) or spins, in the section on ‘Acrobatic Dancing’. The sixth is chakrabhamarī – ‘with one foot in the khand-suchi pose and the other picked up, both combine to perform a perfect wheeling’ (Sarmadee 2003: 110) – a version of which Bose also found in the thirteenth-century Ntyaratnāvalī. Both of these seem comparable to the type of swift spins characteristic of kathak today, although Sarmadee’s translations also contain his own interpretations without it always being clear which is which (Katherine Schofield, personal communication). The Ghunyat al-Munya also originally included information about performing artists, including male dancers called Natwa and female dancers identified as Patur. This data, however, was part of the last four fasls or subchapters, which have been misplaced. Their location and translation would provide an interesting link between the Sanskrit and later material.
Widely available, happily complete and extremely applicable to the search for kathak is the well-known descriptive survey of Akbar’s court, the Ā’īn-i Akbarī of Abul Fazl written in 1593. In the Blochmann English translation (1927), the seventh chapter contains an account of music and dance, providing primarily contextual rather than choreographic details. The chapter moreover includes a long list of the names of communities of performers with a description of the type of entertainment they provided. The list includes singers and dancers (Natwas, Bhand and Kanjari), actors and mimics (Kirtaniya, Bhagatiya, Bhavaya and Baha-rupi), and acrobats who perform feats with ropes (Nats). The list does not include Kathaks, which is in some ways not surprising, as contemporary Sanskrit sources do not include Kathaks either. On the other hand, it again calls into question the supposed migration of the Kathaks, or kathak dance, from the temple to the Muslim court. The evidence in Nartananirāya, written around the same time, indicates that elements of what became kathak choreography were part of court performance but there is no sign of people called Kathaks or Kathakas. Other Persian works, such as the early-sixteenth-century Lahjāt-e-sikandar Shāhi and the seventeenth-century Pārījātak contain sections on dance which may well contribute further information, but have yet to be translated or analysed in any meaningful way (Ahmad 2006).
Dance information in the Tohfat al-Hind written by Mirza Khan in 1675 during the reign of Akbar’s great-grandson, Aurangzeb, on the other hand, has been studied to some extent by Madhu Trivedi (2000). In the Tohfat al-Hind, the tenth section outlines principles of tāl and contains a section on dance (Mohammadi 2006: 51). This description of Nriya Adhyaya, or knowledge of dance, describes four styles: tānd, lās, chain and sudhāng. Defined as a ‘brisk’ and ‘vigorous’ dance, tānd can be connected with tāava, and lās, which is ‘gentle, even and modest’, can similarly be compared with lasya. Tāava and lasya, of course, appear in the Sanskrit literature as far back as the Nāyaśāstra, although lasya is not originally identified as a dance. Chain is less clear. Described as ‘coquettish and pleasing’, it is connected by Mirza Khan to the dance of the triumphant Krishna on the hood of the vanquished serpent Kaliya. This is one of the first apparent links between court dance and Krishna legends, but Trivedi’s connection of chain to the Kaliya gats still danced today is far from convincing (Trivedi 2000: 296). The fourth dance, sudhāng, on the other hand, is ‘a dance form wherein the units of the tāls are performed accompanied with graceful body movements … and expressive and suggestive postures’, a description that once again evokes gat nikās (Trivedi 2012: 174; see also Brown 2003: 73–6).
Chronologically, the next source is the eighteenth-century Muraqqa‘-yi Dehlī, the colourful diary written by Dargah Quli Khan which contains further contextual information (Khan 1989 [1739]). Khan came to Delhi as part of a Deccani delegation to the Mughal imperial court and stayed in the city between 1737 and 1741, witnessing the invasion of the Afghan army of Nadir Shah. His diary documents the daily life which he observed as a cultural outsider and he included details local observers might omit. Along with descriptions of monuments, festivals and various celebrities, Khan included tiny character sketches of noteworthy poets, singers, musicians and dancers. A typically delightful example is: ‘There is another blind musician who uses his belly to create sounds of the dholak and the pakhāvaj, and plays melodious notations. He has invented new rhythms and is an effective accompaniment to the women dancers. Constant striking has turned his belly as black as his luck’ (Khan 1989 [1739]: 95). Khan named and depicted more than 50 performers. His writing includes strong sexual allusions, some of which are homoerotic and purposely mistranslated or glossed over (see Kidwai 2000; I otherwise relied on the 1989 Shekhar and Chenoy translation). There are only 11 performers identified as dancers or described as dancing and Khan did not always clarify what each artist’s specialty was. A few of the performers are linked with groups or performance genres like Domni, Dhadhi, Naqqal or Bhagat-bazan, but there is no mention of Kathaks.
Because he wrote Muraqqa‘-yi Dehlī as a diary, Khan was simply jotting down personal reactions and bits of gossip rather than attempting to provide an accurate report about life in Delhi. One is left, therefore, with tantalizing snippets of information about which it is difficult and perhaps inadvisable to generalize. Furthermore, Khan’s diary is a personal account written during the socially unstable time of Nadir Shah’s invasion of Delhi and some of the characters described seem to be members of a somewhat decadent subculture that cannot be assumed to represent eighteenth-century Mughal cultural life(Brown 2003: 295–6). Nevertheless, there are some details that are useful in the search for kathak. Although the sex of many of the dancers is unclear or mistranslated, those labelled ‘beloved’ or ‘courtesan’ seem to have high economic status and some social power. Khan described most of them as both singing and dancing, and a few as reciting or acting as well. Khyāl is the vocal form he most frequently named, but he also mentioned jangla, tarāna and kabbit (kavitā). There are only three crumbs of choreographic description. Sultana, the coquettish 12-year-old youth, ‘amazed everyone with his … intricate footwork’ (Khan 1989 [1739]: 112), the courtesan Kali Ganga’s dancing is ‘akin to elegant walking’ (116) and another courtesan Zeenat’s dancing is appealing ‘although … like stylized walking’ (117). Footwork, of course, appears in Sagītaratnākara and can be easily associated with kathak. The ‘stylised’ walk is also noteworthy and a comparison may be made with the gatis in Nartananirāya and the gliding cāl found today in gat nikās.
Summary
The Sanskrit and Indo-Persian sources cover a range of language, culture and context, not to mention over 1,500 years of history. The Sanskrit works generally seem to have been written in order to preserve and disseminate the rules of ‘classical’ music, dance and drama. Yet, although early treatises like the Nāyaśāstra show little connection to today’s kathak, later treatises like Sagītaratnākara and Nartananirāya contain descriptions of regional dances which have kathak-like elements. The early Persian works, on the other hand, were written in order to educate the new aristocracy about Indian culture. They contain translations of various Sanskrit sources and add commentary that not only describes more contemporary practice, but also may on further examination support and augment the dance descriptions in the later Sanskrit works. It need hardly be reiterated that there is much that can be done in the translation, analysis and comparison of these texts. Nevertheless, some conclusions are possible with the data at hand. As I stated above, scholars looking for kathak in the treatises have had two goals: the search for priestly storytellers called Kathaks or Kathakas, and the examination of the choreographic descriptions for postures, gestures, movements or dance items which might be connected to today’s kathak. I argue that the quest for the Kathakas is fruitless and based on the Orientalist need for ancient origins, but that the identification of spins, footwork and other choreographic fragments does indeed show some roots of the dances that became kathak.
The search for ancient storytellers has revealed scattered references to people identified in Sanskrit as Kathakas in material ranging from the ancient Mahābhārata to the thirteenth-century Sagītaratnākara. In twentieth-century literature specifically about kathak, this name is left untranslated with the intention of indicating a link to the present-day Kathak community (Hein 1972, Kothari 1989, Narayan 1998). Scholars unconcerned with dance, however, translate the word kathakā as narrator or reciter, and make no explicit link to any present-day performance form (as in Ganguli 1883–96 and Sarmadee 2003). One finds the Kathakas or storytellers most often associated with kings and courts, but surrounded by contrasting company: ascetics and Brahmins in the Ādi parva, singers, rope dancers and fighters in the Anuśāsana parva, a king listening to a love story in Kathāsaritsāgara, and finally bards, scholars and flatterers in Sagītaratnākara. Although one might well search for some similarities between these contexts or attempt to construct some sense of historic continuity linking these references, it seems equally or more logical to understand the term as referring generally to storytellers.
There are, however, other performers mentioned in the treatises and it is they, rather than the Kathakas, who are found performing kathak-like dance movements. The early Sanskrit works seem to favour female dancers, but Sagītaratnākara includes the names of several types of male performers as well. Of interest is the rank of Nata, which is also included in the lost fasl about Natawa from Ghunyat al-Munya. The root of this name is nāya, which refers to the dramatic arts and is most prominent in the title of the Nāyaśāstra. Sarmadee offers more detail about Nata in his ‘Translator’s Supplement’ to Ghunyat al-Munya, describing them as ‘professionals’ who performed ‘acrobatics, rope-dancing, jugglery, always linked with singing and dancing, as forms of popular amusements’ (Sarmadee 2003: 140). The Kathakas are not included in the Ghunyat al-Munya, however, even in the section entitled ‘Concerning Musical Assemblies’, and Sagītaratnākara, which includes both Natas and Kathakas, makes no connection between the two. A connection between Kathakas and Natas should be desirable, because some of the choreography connected with today’s kathak was performed by the Natas rather than the Kathakas. The ‘class-dance’ or marga dance described in the Nāyaśāstra and other early treatises generally has no connection with kathak choreography, although certain items have been imported into it (as in the case of the mudrās). The regional or deśī dances which appear first in Sagītaratnākara include several items – footwork, spins, kavitā – that are present in contemporary kathak. Nartananirāya and the Persian material further reinforces this, and add jakkaī, chain, sudhāng and the stylized walking of the eighteenth-century courtesans in Muraqqa‘-yi Dehlī, all of which also appear related to items in kathak.
Although many of the dance items listed above can be associated with movements in today’s dance, it still seems premature to identify any of these glimpses as direct ancestors to kathak. None are called kathak and none are associated with the storytellers who appear in the manuscripts before 1300. More importantly, the dances themselves are not part of one performance genre, but danced by different people in different contexts and for different purposes. The fundamental mistake in searching for the ancestor of kathak in the treatises is the assumption that the dance of today can be traced back to ancient India if only one looks hard enough. The search for kathak in pre-nineteenth-century Indian Sanskrit and Indo-Persian documents has nevertheless uncovered a number of pertinent and interesting choreographic fragments. These are never called kathak or performed by people called Kathaks and are presented in quite disparate artistic contexts for quite different functions, but some of the roots of today’s dance can indeed be found in these dances from the thirteenth through eighteenth centuries.