Narrative hand-scrolls (pata) have been assembled and painted by the painter-minstrel (patua) communities of Bengal to tell stories for at least two hundred years. Such itinerant bards have traditionally employed the picture sequences to sing well-known stories from the lives of deities (fig. 4.1) and saints, the epics, and more recently to address contemporary social issues and political events.
These scrolls have a distinctive vertical format of framed registers assembled from sheets of paper that are aligned and stitched to the required length for a particular rendition of a narrative. The whole is then backed with thin cotton fabric, usually extricated from worn garments such as saris and dhotis which have the length required to accommodate the images’ continuous vertical format. Nineteenth-century scrolls surviving in museum collections vary greatly in length, ranging from eight panels to over fifty and from five to forty feet, perhaps more given the longest ones are incomplete (fig. 4.2). This length is integral, determining the field for designing the pictorial space, providing the physical parameters for a distinctive visual organisation to the narrative, and undergirding the conventions of handling the object in performance. It also corresponds to the rhyming couplets composed to accompany the pictorial sequence during a performance (figs 4.3–4.6).
Older scrolls now in museum collections have, however, become estranged from the multi-sensorial dynamics between performer, audience, and object, all once vital to their original performances. When rolled up in museum storage vaults or displayed in galleries for viewers to traverse their entire length they are ostensibly stable objects (fig. 4.7). I have previously attempted to juxtapose words of songs with visual registers of scrolls, to imagine a range of relationships between verse and visual evoked in performance. This essay builds on that work to attend to the bodily rituals of handling and touch, which must have mediated viewers’ and performers’ encounters with such objects. Like the aural dimension, these ephemeral practices can only be reconstructed from the variations discernible today. Yet this creative hermeneutic frees us to contemplate the social and performative qualities of objects often in museums, taking into account the lyrics of songs and interaction with the body of the singer, to explore their fundamental performativity and improvisational quality.
The performative skills of each patua are, after all, unique (fig. 4.8). Skill in handling their scrolls in concert with the tempo and rhythm of the song necessarily varies from one performer to the next, one scroll to another, and one session to the next. The coordination of verbal and bodily components with the framed scenes of images can generate distinctive interpretations. A singer may recognise particular visual properties in a sequence of images or consonances between verse and picture, while unfurling the scenes for an audience; bolder or more skilled practitioners may choose to explore these through particular inflections of voice or gesture of hand. Such relationships can turn on the repetition or variation of colour and compositional choices, which may be underscored by the guiding finger, emphasised or subverted by the words sung. Some patua are charismatic entertainers with powerful singing voices who fill performance venues and mesmerise audiences, offering interpretive nuance through skillful manipulation of the lyrics, intonation, rhythm, and tempo. Not all, however, are so overtly dramatic. Some turn the bamboo handles of the scrolls with an elegant turn of the wrist, intensifying the inherent malleability of the objects in generating movement and pauses or even stasis, while displaying a sequence of panels. Others choose to display distinctive propensities in painting their imagery, rather than performance. Indeed, various combinations of performance style can be observed, for example, among the living patua community active in the village of Noya, in Medinipur district of south West Bengal (figs 4.9 and 4.10). Moreover, the occasion for their displays, including setting and time constraints, and the familiarity and affinity of the audience with the performer’s themes, likewise factor in animating the performative features of their scrolls. To complicate matters further, contemporary practice indicates a range of formal and informal collaborations that are not unusual in familial productions in rural communities. Painters borrow (or ‘steal’, depending on the point of view) songs or lines from other composers, the rendition thus varying in nuance and quality among singers. Singers also borrow and use scrolls originally painted by others in their community, equally opening up creative spaces for interpretation and improvisation upon another painter’s visualisation. Each storyteller negotiates the consonance of verbal to visual framework through various bodily strategies. Through such variability in display, a scroll takes on unique incarnations at each unrolling.
In an era before cinema and television, such performances were surely experienced in ways that are difficult to imagine, particularly from our screen-filled and screen-dependent perspectives of the twenty-first century. Yet although patua scrolls have been collected since the early decades of the twentieth century, embraced by Calcutta’s intellectuals and artists, their conditions of display do not seem to have been recorded with much specificity. At this time, the scrolls were incorporated into two interrelated discourses as the interests of the burgeoning field of British ethnology and documentation in the colonies converged with the nationalist quest for an essentially Indian art that was untainted by European influence. Major collections, housed in museums in Calcutta, London, and subsequently Philadelphia, were assembled from this confluence of investments. And something of this colonial taxonomy has spilt over into modern scholarship, which tends to identify these scrolls as performance props rather than works of art in museum displays. Focus has mostly remained on classifying visual narratives, transcribing words of songs, identifying processes of making, and discerning changes in practice against shifting patronage patterns, audience interest, and competition from other media. Or it has followed an anthropological preoccupation with the lifestyles of the living community of makers, their liminal position as craftsmen, and their complex religious and caste affiliations. The increased visibility of women as performers in the past decades, their resourcefulness and success at local competitions and exhibitions, in a few cases leading to international prominence, has also received enthusiastic attention. Yet the engagement of the patua’s body in the physical act of manipulating the length of a scroll to the words of a song has not received the quality of attention lavished on the embodied practice of dancers, martial artists, practitioners of yoga and wrestlers in South Asia. What follows seeks to place the patua’s movements, not merely their lives, back into the picture.
Probing the Ephemeral
The problems inherent in analysing objects as malleable as patua scrolls are formidable, not least because words cannot possibly capture the multi-sensorial and richly affective qualities of the experiences today. If the efforts to translate pictures into words have always been susceptible to multiple impediments, to do so for a multi-media performance act is even more problematic, as anthropologists and folklorists of theatre and performance studies have alerted. To speculate on the past in the absence of much documentation is likewise precarious, inevitably constrained by contemporary observations. Yet, convergence of scholarly interest in the intrinsic material properties of objects as constitutive of their presence or power, and in the sensorial dimensions of experience and embodied practices, offers a lucrative opportunity to heed the performativity and the contingent nature of patua scrolls in practice. I aim to explore such ideas through a single scroll from Medinipur in south-western Bengal, painted in the late nineteenth or at the turn of the twentieth century, now residing in the collection of the Philadelphia Museum of Art. It survives in two fragments—3.65m and 2.90m respectively—and together its halves narrate the life-events of the Hindu god Krishna (fig. 4.1).
This scroll is particularly useful in such a study for several reasons. Firstly, its narrative: by the time this scroll was painted, Krishna worship had gained prominence in the region, and his biography was visualised in many pre-modern art forms including temple sculpture, song, dance, embroidery, and theatre. Together these parallel traditions offer visual, oral, and performative vocabularies and conventions that are integral in assessing the particular choices made by scroll painters to construct their own versions of Krishna’s story. Multiple extant nineteenth-century scrolls on the subject also give us insight into the frequent experimentation undertaken by a particular painter to bring their narratives to life. They allow us to compare the effects with which episodes are visualised or elided—the nature of their juxtapositions either in successive framed registers, in continuous narration within a single frame, or in thematic sequences—as well as how each particular episode is rendered. The Philadelphia scroll, for instance, employs a range of compositional devices for organising the visual narrative along the length of its surface, which could have been manipulated creatively in its display. Most striking is its clustering of episodes along the banks of the dramatic downward flow of the Yamuna River, the site of many of Krishna’s youthful adventures.
Moreover, along with these multiple visual versions, the Krishnalila theme (literally the ‘play of Krishna’, i.e. his exploits) survives in several songs collected early in the twentieth century. Even though these particular ditties were likely not sung to this particular scroll, they still suggest the flavour of verses that could have accompanied such imagery. And these too survive in multiple versions, allowing us to discern how the stock elements comprising these scenes were reworked in various verbal and visual sequences, to create different moods and evoke a range of responses. Their transcriptions offer general insight into the range of possible engagements between the visual and verbal components, strategies that can only be savoured in performance.
Secondly, the Krishna scroll is also revealed as particularly rich when set in conversation with contemporary patua practices, strategies that open up the scroll’s visual possibilities through performative handling. Today, when guided by the words and tempo of their song, a larger scene might allow storytellers a fuller verbal account of a dramatic episode, or for repetition of a refrain for emphasis, or to introduce a pause in the narrative flow. Dwelling on the larger iconic scenes can encourage viewers to quickly grasp them as being significant; to linger on details can imaginatively visualise or insert what is only verbalised. On the other hand, quickly skimming over scenes abbreviated or condensed in a series of recognisable motifs to stand in for well-known episodes, can facilitate establishing particular rhythms and resonances among them through verbal or gestural cues. A quick recounting of the stanzas to the individual episodes, narrated continuously along the width of a panel or across successive registers, may function like a list, to enumerate some of a larger known set, or to suggest episodic equivalence. Conversely, a pithy scene may be elaborated or supplemented by the words of a song, such as the feelings of a specific character or those of others not visualised. A detailed image may be accompanied by a refrain, allowing the viewer to revel in its richness and complexity, or to note divergence from earlier iterations that have since been rolled up but are evoked once again through song. Such performances indicate a range of strategies that energise a scroll and stimulate an array of responses during a multi-media entertainment act, often customising choices to the venue and quality of audience participation.
The patterns of wear along its length suggest that the Philadelphia Krishnalila Pata was well used in performances of just this type and that its imagery might hold within it a palimpsest of the patua’s storytelling movements. As part of a practitioner’s repertoire, its display could have constituted part of a longer performance together with other stories from the regional versions of the epics: the local goddesses such as Chandi and Manasa; local saints such as Chaitanya, Satya Narayan/Satya Pir, and Ghazi Pir; visions of hell (Yama Pata); and many others. From what has been recovered of nineteenth-century practice, based upon oral memory, painter-minstrels traveled with sacks full of scrolls from village to village, where they enticed rural audiences in relatively impromptu performances. Compensation from the audience typically turned on the quality of the entertainment act. Experienced performers likely assessed their audiences and chose their tales and scrolls with an eye toward remuneration. Correspondingly, the verses and refrains may be edited or elaborated as a singer gauges the level of audience engagement. The unrolling of the Krishnalila scroll would thus have been contingent on such conditions as navigated by the individual practitioner, who could potentially have been one of several taking turns in sharing and animating the object over its lifetime.
Unfurling Scenes, Unveiling Miracles
The scroll’s wealth of visual cues for narration and performative potential is perhaps best apprehended if we unfurl it scene by scene alongside song, in the manner of contemporary patua. Its first scene presents the miraculous birth. At the left end of the first register, elevated upon a pedestal, is the now damaged blue figure of Krishna. He would likely have been seated displaying the four-armed form of Vishnu/Narayan. The mother, Devaki, is presented giving birth, lying on a bed, attended by several women. Their raised hands acknowledge the arrival of the baby. One narration suggests a verse counterpart:
And Lord Krishna went to the house of Devaki to be born.
Devaki was lying asleep on the bed
Sitting at the head of the bed
The Lord Krishna showed her a dream.
O mother Devaki Rai sings the patua,
How much longer are you going to sleep
While Lord Govinda wants a place in your womb!
Lying dreaming, the Queen recounts this story:
In my womb, bapu, there is no place for you!
I had six sons and Kamsa killed them all
In grief for my children, I shall weep my life away.
The Lord can’t get a place by any means
So he stole into the mother’s womb in the form of a white fly.
On the eighth lunar day of Bhadra, Govinda [Krishna] was born.
Here Devaki is giving birth to Dev Narayan [Krishna/Vishnu].
This song offers a temporal sequence that is not so readily visible in the visualisation of the episode on extant nineteenth-century scrolls. It narrates at least three successive moments. First, the dream vision is communicated in the voice of the narrator. Devaki’s response is presented as a retelling to her female attendants gathered around her for the birth. The recounting of a miraculous conception returns again to the voice of the narrator, who may well be relying on common knowledge of the wily ways of Krishna among their audience. If an analogous arrangement of verses was sung in conjunction with this panel, the compacted narrative of the visual register would have been expanded verbally, perhaps with a finger guiding viewers from the figure of Krishna to Devaki and her maids, as agency shifts from the one to the next. Such brevity in visualisation could also likely have been abbreviated in the sung verses, customised for a particular audience. The horizontal figure of Devaki could very well stand in for at least three moments, beginning with the announcement from the Divine that he sought to inhabit her body and her negotiation with him, which she is communicating to her attendants during labour. The woman closest to her head may be leaning closer to hear her better. The same supine figure, with her full belly, must surely also be understood as giving birth, with the women supporting her. One attendant is depicted reaching between her feet as if assisting her physically. Not visualised here is the intermediary act, the penetration of her body by the miraculous white fly, a familiar trope in the oral narratives of Krishna’s biography. Scroll performances suggest that a skillful entertainer could have used the single figure to tell successive episodes, using a pointing finger to focus the viewer’s gaze upon the figure for clarification or emphasis.
A second Krishnalila song presents the same basic information, but with variations that could have been significant in performance:
Anybody who uttered the name of Hari in Kamsa Raja’s kingdom
Received beatings on the hands, on the feet and the chest
Where were Basu-Daivakini [Vasudev and Devaki], who had taken the name of Hari?
Taking the form of a white fly Narayan [Krishna/Vishnu] gave a vision:
‘Give shelter to this seed in your womb’.
‘Six sons I bore, Kamsa Raja snatched and killed them
How will another son change my fate?’
First month, second month, the mother cried,
Third month, fifth month, the news spread.
On the tenth day of the tenth month as her pregnancy reached full term
Basumati [Earth goddess] herself served as the midwife, taking Krishna in her lap.
The direct communication between the human and the divine in this iteration, presents the four-armed deity humbly requesting shelter from the devotee. Although Devaki does not resist, she speaks her mind in no uncertain terms, likely knowing that she is ultimately powerless to the presence and will of the omnipotent deity. The intimacy of such negotiations between human and divine, highlighting their mutual dependence, retains the sensibility shared across multiple genres to undergird Krishna worship in the region. Other components of such a performance—again, now lost in museum settings—could have intensified these sung notions even further through intonation, melodic rendition, gestures and facial expressions. The words may have been uttered as the poignant entreaty or angry retort of a woman who has already lost six of her children and refuses the divine for fear of more grief. The range of nuances a storyteller could have brought to the scroll are rich with possibility.
In the next panel, the baby is separated from his mother, now seated on his father’s lap, with the attendants hovering around them. Although the recorded songs do not elaborate on the episode, the painter has presented a wide-eyed Vasudeva prominently at the centre of the panel. His face is rendered in profile, as if turning to look at something beyond. The possibility that yet another divine revelation is being invoked cannot be dismissed. Some viewers may have been familiar with the Bhagavata elaboration of the episode: the father’s amazement at the radiant sight of his son as the absolute godhead; overwhelmed, he recognises the Supreme Lord (Bhagavan) flooding the birthing chamber with divine light. As he bows to acknowledge the divine presence with folded hands and praises him, Krishna assumes his infant form. However, in this pictorial iteration, the rest, their faces depicted in three-quarter view, do not seem to be aware of the vision presented to the father. Successive moments may thereby have been indicated through the presentation of the seated Vasudeva, visualised with profile face, while his body is frontal, with the baby on his lap. Viewers may also have filled in what is neither visualised nor verbalised: that at this moment the shackles fall, the doors of the prison open, and Vasudeva is invited to remove the baby from Kamsa’s stronghold.
The narrative proceeds to the left. The baby is smuggled away by the father, the snake’s hood protecting the vulnerable, naked divine. The two scenes are differentiated by the figures turning in opposite directions rather than any dividing frame. The repeated figure of the blue-skinned baby signals the continuity of the narrative, which proceeds from right to left. Next, Vasudeva carries the baby to the river’s edge. The baby slips out of his hands momentarily and floats upon a lotus leaf on the turbulent river. To the right, he brings the boy to his friend Nanda’s home in the village of Gokul and exchanges the blue-skinned boy for the girl-child born to his friend’s second wife Yashoda. Implicit here is the shared knowledge between the performer and audience that the boy who slipped out of his father’s hands returns to the agitated father.
The painter has thereby chosen to vary modes of visual narration from the first to the second panel, and reversed direction thereafter. Indeed, the appreciation of such juxtapositions may well have constituted part of the pleasures of the performance for a discerning audience. If viewers were already familiar with the general storyline, they would have been able to savour and possibly even seek out the distinctive nuance offered in the particular choices made by the singer to interact with the visual imagery.
In juxtaposing the verses that correspond to the visualisation of particular scenes on the scroll, the possibility of various interactions between them emerges. Some episodes condensed in the visual narrative may have been privileged for elaboration in song, elucidating the extraordinary nature of the events described. Others are mentioned only briefly or elided entirely in the song, relying for the most part on audience fore-knowledge. The song elaborates upon the miraculous qualities of the journey that is visualised: the play of the divine and the compassion of a benevolent god who forgives errors and reveals himself to the devoted father:
Immediately after its birth, Vasudeva took the child in his arms:
Vasuki the snake from Patal made a canopy!!
Slowly they walked down to the Jamuna.
The waves of Kalindi, the river of illusion,
grew bigger and bigger –
Lord Narayan [Krishna] slipped out of Vasudeva’s arms into the river:
Weaving a web of illusion he began to float on a lotus leaf.
Not getting milk from its mother, the child chews its thumb,
While Vasudeva feels about for him with his hands.
‘Alas what has happened, because of my cursed fate –
All these troubles at my ages only to splash about!’
Dayal [Kindly] Krishna seeing the weeping of Vasudeva,
Presses two hands on the water and climbs into his arms.
These verses seem to assume the audience’s familiarity with various details of the tale, including the fact that it was a stormy, dark, moonless night, making the journey a particularly dangerous one. However, natural forces could also cooperate to acknowledge the presence of the divine: if the snake offered an umbrella, a fox led the way, and the turbulent river parted to make way for the baby to be transported to safety. The audience surely also knew that it was to protect the divine baby from the wrath of the local overlord Kamsa. The verse iterations cited indicate that he killed each of Devaki’s children to avert threats to his right to rule over Mathura, as previously predicted.
The composition and colour choices in this scroll also emphasise the miraculous nature of the events. The blue-skinned baby floats upon a golden mandorla, itself afloat on a river which meanders downstream, dynamically rupturing the horizontal arrangement of several registers to underscore the verticality of the scroll. Just below, the yellow-skinned female, who is to be exchanged, floats upon the blue waters in a reversal of colours, as if to draw attention to the roles and fates of these two infants. However, current practice suggests that scrolls are hardly ever viewed opened up in their entirety. As a song unfolds, a patua unwinds to the appropriate scenes, with the help of the bamboo handles at either end; viewers were likely huddled in groups to see the images turning on the bamboo handles. Thus, while the two images could have been appreciated in comparison at the same time, they may have been revealed sequentially with the progression of the song and the tempo of the unfurling registers. Such choices would suggest different viewing experiences, the latter relying on memory for any such comparison.
Another song indicates a different verbal strategy to enhance the supernatural dimension of the story, emphasising the celestial entourage provided for the divine child:
Vasudev proceeded to Nandalaya [home of Nanda Ghosh] to hide the baby.
Seeing Krishna the Yamuna River spilled over her banks.
Bhagavati [the goddess, Durga] took the shape of a fox to escort them in crossing the river.
‘Ten months, ten days having spent in his mother’s womb
The Lord bathes in my womb, so fortunate am I’.
It is possible that these extraordinary events were so familiar to audiences that such fresh verbal and visual interpretive twists renewed awe and wonder (vismaya).
On the right third of the panel, Krishna is delivered to the family of Nanda for safekeeping, and the next register presents Vasudev returning to his wife with Nanda’s baby girl to the left. To the right, Kamsa kills her by hurling her upon a stone on the river’s edge. But she escapes, here flying upstream, revealing her true form as the golden-hued goddess. Her message is suggested in the words of the following song:
‘What have you gained by killing me, O hero?
He who will kill you will be born in Gokul’.
Following this prophecy, Kamsa, in a moment of remorse, seeks forgiveness from his sister and brother-in-law, which they grant. However, he has a change of heart and decides to seek out the baby who was to destroy him. The ensuing panels are devoted to the well-known exploits of the youthful Krishna, vanquishing the series of demons Kamsa dispatched to the village of Gokul in response to the prediction. They include on the right Putana, the ogress who tries to poison the boy while breastfeeding, followed by the crane (Bakasura), the horse (Keshi), the cart demon (Sakata), and elephant (Pramila) in the next series of registers. These scenes are stock types, presenting the episodes in a terse synoptic fashion that features only a single moment in each: that of Krishna accomplishing the annihilation of demonic forms to restore order. A comparison with other extant scrolls indicates that these compositions are used by artists in various sequences along the length of the river to different effect (compare, for instance, with figs 4.11 and 4.12).
The riverbank is also where Krishna spent much of his youth enjoying the camaraderie of the cowherd boys (sakhas, gopas), and the attention of the local women (sakhis, gopis). The handsome adolescent’s dalliances with the cowherd women, and his beloved Radha in particular, are visualised in three scenes. They are part of a larger corpus of episodes introduced earlier in Bengali poetry and temple imagery to imagine his relationship with Radha, the young woman who stole his heart. They may have offered beholders the opportunity to remember and re-experience the emotional intensity that the divine lovers had themselves felt.
In the first episode, posing as tax collector, Krishna waylays the women, demanding payment to cross the river (danalila):
Pay the toll and get on the boat Radhe Vinodini (Radha, Bestower of Pleasure)
To take all of your companions I will charge an anna each
To take Radha I will have a golden earring.
‘Take the gold, take my sari, I can give you everything
I cannot cross the river on my feet’.
Hearing this Krishna escorted her across.
One by one the gopis went to Mathura.
On the opposite bank of the river, an earlier episode in the life of the baby Krishna is presented. He is shown high up in a tree that is splitting apart at the bottom, releasing the twins Jamal and Arjuna from the pair of trees in which they had been imprisoned. Again, not depicted, and possibly not even elaborated in song, are the events leading up to the climactic moment that is privileged for visualisation. Viewers likely knew that the child had been tied to a heavy mortar by his mother by way of punishment for his mischievous misdemeanours, and that Krishna was determined to release himself by crawling between the two trees, so that the mortar would be left behind as the rope gave way. In the process, the trees collapsed, releasing the sons of Kubera.
The visual arrangement of episodes again departs from a strictly chronological unfolding of his biography, as two scenes of the child milking cows in his father’s shed and stealing butter from his mother’s churn, follow from the trysts of the lovers. Such flexibility in arrangement of the visual sequence is facilitated by a number of factors. First, the narratives are very popular and the individual episodes have acquired discrete lives in image and text within their synoptic visual iterations. Second, in performance the sung verses can create a continuity through rhyme, rhythm, and repetition, which may appear abrupt or even confusing if the visual were to be experienced in the absence of the verbal component. The songs recorded by Dutt reveal that words of the songs employ various strategies to effect transitions from one episode to the next. Indeed, multiple disjunctive temporalities emerging between visual and verbal sequences, and those anticipated from prior knowledge, could potentially have enhanced audience participation.
The next panel offers another unusual pairing of episodes that may be distinctive to the genre. The words of a song would suggest it could have been navigated from right to left. On the right half of the next panel, Krishna carries Radha’s dairy goods to market (bharakhanda). When he shows up in the garb of a porter in the woods, she has him carry her dairy goods to the market, now promising her favours in exchange. A patua song outlines the negotiation between them:
I’ve never carried loads, I’m Hari, Lord of the Universe,
The unruly handle burns my shoulder.
Radha, says to the Lord, You’ve taken my money, now you’re obligated to do so.
Now why do you say you can’t bear the load, Dinanath [Lord of the Poor and Helpless]?
Removing the load, Vanamali [Krishna] sat.
Hiding her face, Chandravali [Radha] smiled.
However, she manages to give him the slip after the market, choosing to return home on a different path. The two episodes of toll collection and porter services visualise the power dynamic in the lovers’ affair and the complexity of emotions they experience. Exploring intimate exchanges, the verses also remind the audience that the absolute divine had chosen to experience Radha’s love alongside performing the more conventional salvific functions of an avatar. She, however, can only see a troublesome young cowherd youth who shamelessly and relentlessly chases her. She seems to have the upper hand, for the moment.
This episode, in which she is in control of the lovesick youth, is placed across from the display of one of Krishna’s best-known cosmic feats. If the youthful lover protests at the menial task of carrying dairy goods for his beloved, he certainly takes full responsibility for his community in lifting Mount Govardhana with supreme ease, barely raising one finger to create an umbrella and shelter his village from the onslaught. Challenged by Indra, who showers rain relentlessly, Krishna reveals his true form as the divine savior. Both episodes are immensely popular and the artist may well be relying on that foreknowledge to create a distinctive juxtaposition and interpretation. Encouraged by these visual resonances and dissonances of form and composition, and perhaps alerted by similar balancing of power in the songs of older poets such as Chandidas and Jayadeva, the viewer may have appreciated the patua’s inventiveness. If the visual invites comparison of these two episodes of load-bearing, a song offers further nuance:
‘I can’t carry this, Radha, the load is too heavy’.
‘Why Krishna did you eat the yogurt, your payment?
You have to take the load to Mathura’.
Krishna recalled Indra:
Indra brought rain, Pavan brought wind,
Crossing the river of illusion, he reached the shore.
The raising of the mountain, a feat of extraordinary force and one that came to take on a ritual life of its own, is referenced only as a recollection in this song. It not only reminds viewers of Krishna’s true strength in his cosmic form, but is presented as motivation for a young man whose ego had clearly been hurt at being treated as a lowly porter. The dual dimensions of Krishna as absolute divine and simultaneously utterly mortal are simultaneously drawn out: a besotted youth chafing inwardly as he chases his love, overcoming a humiliating experience through exertion of mental images and exercises. The intimacy of the emotional dimension is valued deeply in this religious tradition as avenues to engage in exploring an intensely personal relationship with the deity. The reward for such dedication is the remembrance of the illusory nature of human existence, one that Krishna here offers as role model and promise to the audience.
Naukavilas (pleasures of the boat), Krishna’s boat ride with Radha and her companions across the Yamuna River, follows the toll and porter roles on the right half of the next register. Here Krishna raises his oar at the far end of the peacock-prowed boat (mayurpankhi). He balances easily at a rakish slant, leaning over a woman, likely Radha, whom he had sought out. The composition shares many elements with terracotta, embroidery, and print versions, which could have invoked appreciation of the painter’s rendition, although the compositional flourish in filling the curve of the river with the shape of the boat is not to be found in embroidered or printed vignettes, which do not offer a series of episodes in vertical succession. A patua’s song could have recounted the well-known tale of Krishna, disguised as a boatman at the river’s edge, ferrying the women across on condition of Radha’s affections as payment, as in this version recorded by Gurusaday Dutt:
‘Pay the toll and get on the boat Radhe Vinodini (Radha, Bestower of Pleasure)
To take all of your companions I will charge an anna each
To take Radha I will have a golden earring’.
‘Take the gold, take my sari, I can give you everything,
I cannot cross the river on my feet’.
Hearing this Krishna escorted her across.
One by one the gopis went to Mathura.
Many versions of the episode exploring shades of emotional entanglement continue to be sung in temple courtyards and home shrines today. If this patua presents a helpless Radha giving in to the rake’s mercenary demands, the audience likely knew some of the other interpretations read into the episode through the centuries in various media.
Following the episodes clustered along the banks of the river, a set of three panels is devoted to Krishna’s bovine encounters. Cows were not only the mainstay of the rural community of Krishna’s childhood, but also relevant to the predominantly agrarian economies of the painter communities and their audiences. At left, Krishna is milking a cow, possibly stealing milk as he does with the butter from his mother’s churn on the right, as a verse suggests:
Dark Krishna happily milks the white cow.
The milk overflows, so Krishna opens his mouth [to drink directly from the udder]
The butter thief episode (nonichora), as narrated by patuas, likely assumes their viewers’ knowledge of one of the best-known miracles performed by the divine boy: the wondrous vision he offers his mother when she scolds him for dipping into her churn, after which she peers into his open mouth and glimpses the entire cosmos, including herself and her home in Braj. Viewers surely knew that such a revelation of ultimate truth, offering an omniscient viewpoint unavailable to human eyes, must have bewildered and overwhelmed her. If the stock image is shared with many other genres, the extant verses also do not seem to elaborate:
And Shandashur [bull demon] was killed by Damodar in the cowshed.
Tying the calves in the forest, he milked the cows.
Mother Josoda [Yashoda] of good fortune was churning the milk
The two brothers Krishna and Balaram stole the butter;
The Lord drank the skimmed milk and lay down to sleep…
This song moves from one episode about the cowherd life of Krishna to another, but in reverse order to the scroll. Instead, the following register visualises the defeat of the bull. Krishna is in the act of impaling him by the horns as the bull charges at him. It is presented with another feat, depicting Krishna atop a gigantic demon figure. This may be the defeat of the demon in the form of the ass Dhenukasura, who guarded the palm forest and denied the local population of the delicious fruit or another presentation of the demoness Putana.
The third scene relating to the bovine life of the village and Krishna’s interventions chosen by this artist, is the tale of Brahma stealing the cows from the cowherd boys to test Krishna. Again, only the climactic moment of Brahma’s realisation that the boy was indeed the absolute divine is depicted, assuming general familiarity with the events leading up to it. Here Brahma returns the cows he had kidnapped on recognising Krishna’s omniscience.
A large panel, double the size of the previous ones, displays the popular tale of Krishna stealing the clothes of the women while they bathe in the river. Here he is presented walking along a tree branch, gazing below at women who plead with raised hands. Pink articles of clothing dangle from the spreading branches of the kadamba tree, along with its globular blossoms. A song offers some sense of the verbal accompaniment possible:
Stealing all the sakhis’ clothes Govinda
Is sitting in the kadam tree, playing on his bamboo flute:
‘O moon-faced sakhis, now listen to me
Lift up both your hands and ask for your clothes back!’
Understanding their state of mind, Krishna
Threw back their clothes into one pair of hands after another.
If the image suggests the adolescent’s desire, the songs divert attention to the women’s state of mind, transforming the tale into one of divine compassion.
After the telling of the scroll’s tale is accomplished, the patua typically rolls up the scroll to return it to the pile of dormant ones. At this point, if the audience is hooked, they can make specific requests and offer money as they would to street singers. However, if the performance is deemed indifferent, the crowd thins and peters out.
The relatively small size of all but the final surviving register of the Philadelphia scroll suggests that more than one register was probably revealed at a time. The span of the scroll that an individual can display at a time is also dependent on the length of an arm. Among living practitioners today from the Noya community, it is typical to hold the scroll up with one hand to reveal the scenes from the top of the object, while the other keeps the rolled-up portion at the bottom. In observing Rani Chitrakar, one of the most successful contemporary women singers of Noya, I came to appreciate her dexterity in manipulating the object (fig. 4.3). She is keenly aware of how far she can stretch her arms, and how she can use her entire body to facilitate the display of the scroll. For example, she uses a bent knee to create additional support, from behind, for the display of the expanse of the scroll between her hands. It allows her to control the length of paper, the angle at which she presents the image, and frees up a hand to use as a pointer. She frequently casts a sidelong glance to ensure that the segment exposed is as intended, and that the angle allows for ease of viewing. Her skills in presentation are not unlike of those of a television weather reporter, coordinating between the screen and the viewer, using sideways movements of the hands, head, and eyes, while maintaining a frontality that consistently engages the viewer. It also became apparent during my fieldwork in Noya that performers demonstrate varying degrees of skill and ease in turning their wrist and using the thumb, for example, to roll up a scene (figs. 4.4–4.6). Such contingencies have several consequences: some slow down the tempo of the song to coordinate the visual to verbal components, while others deploy the refrain strategically to create a pause to roll up a section.
Such bodily engagements with scrolls suggest ways to imagine the Philadelphia scroll fragment opened up in sets of scenes that cohere around particular subthemes, which may be conceptual or visual. For example, the birth and transfer scenes may have been opened up together, while the river scenes—rupturing the strict division of registers, physically directing our gaze downward, much like an arrow—could have been opened up as a set if the hands of the performer were long enough and the skill of using the knee to support the object from the back were employed successfully. It is tempting to imagine a deft patua exploiting the painted cues of the scroll’s visual organisation, to present a large scene of the theft of the clothes alone in dramatic contrast with the dense clustering along the length of the river, or pairing it with another large scene.
Variations surely emerged in the handling of the scroll from one entertainment act to another. The first three scenes could probably easily have been displayed comfortably between the hands and feet of a singer. Contemporary performances indicate that the bard does not necessarily adhere strictly to the painted frames of scenes. Instead, by leaving part of a previous scene visible as she moves onto the next, or exposing a fraction of the next scene, a particular iteration can momentarily affect a visual continuity that can either become consonant with the verbal transitions of the song, or one that is distinct from both verbal and visual cues. Such partial unveilings could potentially operate like a movie trailer or a preview of the next episode at the end of a television show, encouraging the audience not to switch channels. In this case, it could serve to engage the viewers, deter walking away, thereby operating as a strategy for securing remuneration. Regardless of whether the particular expanse exposed in rolling and unfolding is intended or appreciated, it could potentially have introduced greater complexity.
A host of other contingencies emerged in my observation of contemporary patuas. In an impromptu performance Monimala Chitrakar, for example, employed the elevated threshold of a balcony to prop her scroll from behind (fig. 4.13). Likewise, the only time I observed two patuas working in tandem, even greater options emerged than is typically assumed for a single singer. Karuna Chitrakar assisted her husband in unrolling to the timing of his song as he is now blind. Such pragmatic and creative combinations in making use of specific skills and tools today encourages us to be aware that many others cannot possibly be recovered.
Attention to medium and its physical properties thus urges us to take into account the quality of presence that a narrative acquires in particular material incarnations, in the triangulation with entertainer and audience occupying distinctive sites, and in shaping cultural domains at specific historical junctures. The ephemerality of patua scrolls in performance, despite the methodological precariousness, allows for greater awareness of how material properties shape perception, as well as bodily engagements. In turn, opening up such possibilities alerts us to the ambiguities inhering in the ever-changing entanglements between the various media through which distinctive regional images assume visible form.
Pika Ghosh is Visiting Associate Professor of South Asian Religion and Visual Culture at Haverford College. Her research interests range from early modern temple architecture to repurposed textiles and their embroidery in colonial India, terracotta sculpture in the religious visual culture of eastern India, and painted paper handscrolls as a form of performance. These projects span time, class, and media from Mughal to the colonial period and independence, and from domestic, intimate, everyday concerns to temple and court ritual. A deep commitment to the object and its materiality as the source of multiple voices that engage in historical and cultural changes runs through them. Her first book, Temple to Love (Indiana University Press, 2005), examining the role of a distinctive regional architectural form in framing devotional practice, received the inaugural Edward C. Dimock Prize for First Book in the Humanities from the American Institute of Indian Studies. Her second monograph, Making Kantha, Making Home (University of Washington Press, 2020), investigates embroidered textiles to recover women’s voices from colonial Bengal. This project builds on research on kantha for the 2009 exhibition at the Philadelphia Museum of Art, which situated the fabrics within domestic networks, entangled with memories, perceptions, sensorial resonance and emotional experience. The publication received the College Art Association’s Alfred H. Barr Jr. Award for museum scholarship.
 The distinction between social and religious subject matter is one that performers emphasise today and it is likely related to the nature of commissions received. Although the extant collections of nineteenth and early twentieth-century scrolls privilege the narratives of gods and goddesses, this may be contingent upon the collecting interests of the early twentieth century and the colonial and early nationalist investment in devotion and spirituality.
 There is also a wide range in the quality of materials and labour invested in their assembly. Some were created from better quality paper and painted with great care; lavished with exquisite details that indicate awareness of the stylistic, compositional and narrative conventions of painting evinced in the older manuscripts shared across north India. See, for example, the Victoria and Albert Museum’s Ramayana Scroll (I.S. 105-1-39 1955). I have examined this scroll in ‘Scenes from the Rama Epic’, in Forrest McGill (ed.), The Rama Epic: Hero, Heroine, Ally, Foe (San Francisco: Asian Art Museum, 2016), pp. 13–17. Other scrolls are put together more quickly. Swiftly painted broad brushstrokes and far less attention to detail create a markedly difference in style as seen in a second scroll devoted to the epic in the same collection (I.S. 107-1-7 1955). On this scroll, see my study ‘A Bengali Ramayana Scroll in the Victoria and Albert Museum Collection: A Reappraisal of Content’, South Asian Studies 19:1 (2003), pp. 157–167. Some scrolls are assembled from repurposed newsprint and other cheap paper extracted from railway timetables, the text sometimes visible through the layers of paint. The Philadelphia Museum of Art collection offers several examples of such recycled paper.
 Whether the visual skeletal structure is worked out first or the song composed before the images are conceptualised varies among the living practitioners today and is contingent on several factors, among the most significant being their personal skills.
 I have conducted fieldwork since 1995 both in the patua village of Noya in Medinipur district and among performers at various venues in fairs at Shantiniketan and Kolkata, as well as at museums in Delhi, Chennai, London, and Philadelphia.
 They can range from more contextualising commentary on the depicted image, to an elaboration of details, including physical ones, that have not been delineated visually, and can evoke the affective dimensions of the narrative by raising questions about the moral, ethical or social justice issues in a particular event or state of affairs. I have discussed this elsewhere in: ‘Story of a Storyteller’s Scroll’, RES: Anthropology and Aesthetics 37 (2000): pp. 166–185; ‘Unrolling a Narrative Scroll: Artistic Practice and Identity in Late Nineteenth-Century Bengal’, Journal of Asian Studies 62:3 (August 2003): pp. 835–871; ‘A Bengali Ramayana Scroll in the Victoria and Albert Museum Collection: A Reappraisal of Content’, South Asian Studies 19:1 (2003): pp. 157–167; ‘Rasalila Remixed: What Gets to Stay?’, in Pallabi Chakravorty and Nilanjana Dasgupta (eds), Dance Matters Too (New Delhi: Routledge, 2018), pp. 137–167.
 I draw upon the proposition offered by a range of influential scholars across several disciplines. Arjun Appadurai, drew attention to the ways in which objects were successfully moved and recontextualised. While Appadurai attended to the dimension of commodification and decommodification, Igor Kopytoff’s essay opened up the question of the mutability of artworks, artifacts, and objects. Arjun Appadurai, ‘Introduction’, in Arjun Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective (Cambridge: Cambridge University Press, 1986); Igor Kopytoff, ‘The Cultural Biography of Things: Commoditisation as Process’, in Appadurai, The Social Life of Things, pp. 64–91. The heightened attentiveness to the malleability of objects also reopened the question of agency, particularly by Alfred Gell, who called for attention to the latent intentionalities of creators inhering in the very materiality of the works, their concomitant capacity to evoke emotional responses, and hence constitute social spheres. Alfred Gell, Art and Agency: A New Anthropological Theory (Oxford: Oxford University Press, 1998); Alfred Gell, ‘The Technology of Enchantment and the Enchantment of Technology’, in Jeremy Coote and Anthony Shelton (eds), Anthropology, Art, and Aesthetics (Oxford: Clarendon Press, 1992), pp. 40–67. Bruno Latour’s exhortation to seek out, through empirical observation, the dynamics in the jostling of objects, plants, theories, texts and people, to uncover the specifics of the relationships rather than a priori assumptions, has been highly influential. Bruno Latour, Reassembling the Social: An Introduction to Actor-Network theory (Oxford University Press, 2005). Hans Belting redirected focus to the charge, or animation, in his terms, of the object in the triangulation between image, its material incarnation and the viewer. Hans Belting, Likeness and Presence: A History of the Image Before the Era of Art (Chicago: University of Chicago Press, 1994), and Hans Belting and Thomas Dunlap, An Anthropology of Images: Picture, Medium, Body (Princeton and Oxford: Princeton University Press, 2014).
 Although the scholarly literature has insisted that scenes are unveiled one after another, I have not observed such precision; in addition, video recordings equally offer evidence to the contrary. The lack of consistency in size of registers relative to one other and the variability in the length of a performer’s extended arm and fingers holding and rolling and unrolling the object, inevitably belies such fastidiousness.
 Roma Chatterji recognises this variability and usefully explores meta-textual relationships concretised in specific narratives, such as the newer Laden Pata (Osama Bin Laden Scrolls or 9/11 Scrolls), which she locates with the established stories of pirs, as part of shared worlds of storytelling familiar to some traditional audiences. Roma Chatterji, Speaking With Pictures: Folk Art and the Narrative Traditions in India (London: Routledge, 2016), p. 90.
 Scrolls were collected by colonial administrative figures such as Gurusaday Datta and J. C. French, connoisseurs like Ajit Ghose, and artists including the Tagores, Nandalal Bose, and Jamini Roy in the early decades of the twentieth century, as well as the art historian Stella Kramrisch. Important historic collections include the Asutosh Museum’s material, collected under the supervision of its first director, Deva Prasad Ghosh, and the Gurusaday Museum’s collection, acquired by Datta, in Kolkata. Outside India, the Victoria and Albert Museum and the British Museum have some of the most exquisitely rendered scrolls. Stella Kramrisch brought her patas to the Philadelphia Museum of Art. Aside from these better-known collections, patua families have continued to collect casually, occasionally selling their material to museums and connoisseurs. Dutt argued that because the ‘folk’ arts were authentically Bengali, they provided the answer to the quest for a nationalistic art, ‘The Indigenous Paintings of Bengal’, Journal of the Indian Society of Oriental Art 1(1933): pp. 18–25.
 Studies on style and establishment of regional stylistic groupings include Kavita Singh, ‘The Pictures of Showmen’, (Ph.D. Diss., University of Chandigarh, 1995) and T. Richard Blurton, ‘The “Murshidabad” Pats of Bengal’, in Jyotindra Jain (ed.), Picture Showmen: Insights into the Narrative Tradition in Indian Art (Mumbai: Marg, 1998), pp. 42–55. In my own fieldwork among living practitioners, many attributed the decline in the practice to the competition from newer media, especially cinema and television in the second half of the twentieth century. Resourceful patuas have long compensated for the dwindling livelihoods with an astonishing array of innovations. Intervention from local governmental agencies, and urban and cosmopolitan investment in ‘folk’ art, equally contributed to transformations in the practice. These features have attracted much attention in recent decades, from both urban elites and academics in various countries. In turn, these forces have participated in stimulating a revival and renewal, inevitably bringing with it significant transformations, including the cultural value and status of both objects and practitioners. On recent transformations in painting style, subject matter, materials, venues, and patronage patterns in the living tradition, see T. Richard Blurton, ‘Continuity and Change in the tradition of Bengali Pata-painting’, in Anna Libera Dallapiccola (ed.), Christine Walter-Mendy and Stephanie Zingel-Ave Lallemant (collabs.), Shastric Traditions in Indian Arts, vol. 1, (Weisbaden: Franz Steiner Verlag, 1989), pp. 425–451; Kavita Singh, ‘Stylistic Differences and Narrative Choices in Bengal Pata Painting’, Journal of Art and Ideas 27–28 (1995): pp. 91–104; Kavita Singh, ‘To Show, To See, To Tell, To Know: Patuas, Bhopas, and their Audiences’, in Jain (ed.), Picture Showmen, pp. 100–115; Singh, ‘What’s New in Pata Painting’, in Anna Libera Dallapicola (ed.), Indian Painting: The Lesser-Known Traditions (New Delhi: Niyogi Books, 2011), pp. 63–79; Alliance Francaise and Crafts Council of West Bengal, Patua Art: Development of Scroll Paintings of Bengal Commemorating the Bicentenary of the French Revolution (Calcutta: Peerless, n.d.); Sankar Sengupta, ‘The Patas of Bengal in General and Secular Patas in Particular: A Study of Classification and Dating’, in Sankar Sengupta (ed.) The Patas and Patuas of Bengal (Calcutta: Indian Publications, 1973). Roma Chatterji offers a useful case study on shared and divergent approaches to a particularly popular narrative, 9/11 (Laden Pata), by comparing the visual and verbal iterations of several painter-performers from a related community of living practitioners, Speaking With Pictures. Such transformation had been evinced in the emergence of new genres in the nineteenth century, such as Kalighat painting, from the adaptation of traditional scroll paintings to the new interest in souvenirs as Bholanath Bhattacharya noted in ‘The Evolution of the Kalighat Style and the Occupational Mobility of the Patuas. A Sample Survey’, in Sankar Sengupta (ed.), The Patas and Patuas of Bengal (Calcutta: Indian Publications, 1973). More recently, Roma Chatterji has attended to the transformations stimulated by intermedial experiments (entextualisation in her terms) including animation and graphic novels, Speaking With Pictures. See also Bhaskar Mukhopadhyay, ‘Dream Kitsch – Folk Art, Indigenous Media and 9/11: The Work of Pat in the Era of Electronic Transmission’, Journal of Material Culture 13 (2008): pp. 5–34.
 Scholars such as Binoy Bhattacharjee have examined the self-presentation of the community in various performance contexts and in their personal lives, responding in a general fashion to the assumption of essentialised Hindu and Muslim identities, since colonial census-taking practices delineated these categories for classifying peoples. See Bhattacharjee, ‘The Patuas: A Study on Islamisation’, in Sankar Sengupta (ed.), The Patas and Patuas of Bengal (Calcutta: Indian Publications, 1973), pp. 95–100; Binoy Bhattacharjee, Cultural Oscillation: A Study on Patua Culture (Calcutta: Naya Prokash, 1980).
 These include Malini Bhattacharya, Women Potuas of Medinipur: Exercises in Self-Empowerment (Calcutta: School of Women’s Studies, Jadavpur University, 2004); Akos Ostor and Lina Fruzzetti, Singing Pictures: The Art and Performance of Naya’s Women Painters (Lisbon: Museu Nacional de Etnologia, 2007); Frank J. Korom, Village of Painters: Narrative Scrolls from West Bengal (Santa Fe: Museum of New Mexico Press, 2006); Singh, ‘What’s New in Pata Painting’.
 See for example, Phillip B. Zarrilli’s work on the martial art forms of the Kerala region including, ‘Repositioning the Body, Practice, Power, and Self in an Indian Martial Art’, in Carol A. Breckenridge (ed.), Consuming Modernity: Public Culture in A South Asian World (Minneapolis and London: University of Minnesota Press, 1995); and Pallabi Chakravorty, Bells of Change (Calcutta: Seagull Books, 2008). See also Joseph Alter, Yoga in Modern India: The Body Between Philosophy and Science (Princeton: Princeton University Press, 2005).
 The dilemma of limiting the possibilities, and projecting our perceptions upon the past, however, is one we have come to acknowledge across disciplines. I take the lead offered in the methodological interventions by scholars such as Irene Winter, Stephen Greenblatt and many others to animating the past, with awareness of our own predispositions and proclivities as much as possible. Pithy versions of these methodological forays are available in such essays as Winter, ‘Idols of the King: Consecrated Images of the King in Ancient Mesopotamia’, Journal of Ritual Studies 6:1 (1992): pp. 13–42, and Greenblatt, ‘The Circulation of Social Energy’, in Nicholas B. Dirks, Geoff Eley, and Sherry B. Ortner (eds.) Culture/Power/History: A Reader in Contemporary Social Theory (Princeton: Princeton University Press 1993), pp. 504–519.
 The scroll currently exists in two pieces (Accession No. 1994-148-548a, b), acquired by the distinguished historian of Indian art, Stella Kramrisch, while she was working in India from 1922 to 1950. They came with her to Philadelphia, and were bequeathed to the Philadelphia Museum of Art, where she served as curator until her death in 1992.
 With the development of devotion to Krishna, remembering the stories of Krishna’s life became significant in a range of ritual practices as a way to gain access to the divine. The central text developing Krishna’s biography, The Bhagavata Purana, which came to be regarded as scripture, advocated remembrance of Krishna’s life-stories as a path to salvation: ‘The stories of Lord Krsna [Krishna] are meritorious, sweet, ever-interesting, delightful and they wipe out the sins of the world. What person who knows the importance of listening to them, will be sated at hearing them’. Bhagavata Purana 10.51.20. The version used here is Ganesh Vasudeo Tagare (ed.), Bhagavata Purana, translated and annotated in Ancient Indian Tradition and Mythology, vol. 10 (Reprint, Delhi: Motilal Banarsidass Publishers, 1994).
 Oral performance of song and recitation has been described in courtly venues and temple courtyards. The Bengali mystic Chaitanya, for example, sang devotional songs called padavali kirtan that describe the better-known episodes, and led his followers in ecstatic dance processions through the streets of his hometown Navadwip; this practice became a distinguishing characteristic of the religious community that cohered around him. Single episodes from Krishna’s life were elaborated into dramatic performances. The defeat of Kaliya (Kaliya Daman), the tale of the devotee Akrura (Akrura Sambad), Radha’s grief (Rai Unmadini), and Radha’s Dream (Swapna Vilas) became the subject of panchali and jatra, plays performed at village gatherings. And vice versa, these episodes were stitched together in different combinations to make various points during storytelling. These expressive genres share a common base of narrative elements with the region’s written ones, and also verbal formulae among the oral genres. A storyteller could draw on them to constitute his/her narrative and delineate characters, while injecting his/her personal reflections, and relying on the audience’s familiarity with the themes to create particular inflections.
 Not surprisingly, episodes from Krishna’s biography are clustered to different effect. While some emphasise a range of popular episodes from his birth, childhood to adolescence, others focus primarily on his love affair with Radha. The Victoria and Albert Museum’s Krishnalila Pata (IS 1955-108), for example, offers useful comparison in its choice of several episodes that are omitted in the Philadelphia scroll, as well as arrangement of scenes that they share.
 The object in its current state is incomplete, probably the central section of a longer scroll. It may have been preceded by an iconic oversized register invoking the deity, following the practice shared by several nineteenth-century scrolls devoted to this narrative, including the Victoria and Albert Museum scroll (IS 109-1955). This convention continues into the present, for example in Medinipur scrolls devoted to goddesses. Physical examination of the bottom edge also indicates that some scenes are missing. Complete Krishnalila scrolls that incorporate most of the scenes visualised on the Philadelphia scroll, often end with a series of large images of Krishna quelling the snake Kaliya and his dance with the gopis (rasalila). Scrolls dedicated to Vishnu/Krishna can also incorporate images of the other incarnations of Vishnu/Krishna, as in the case of the Victoria and Albert Museum scroll (IS 108-1955).
 The painter takes advantage of the vertical format to enhance the downward movement of the narrative by cascading episodes down the length of the river. The river, viewers surely knew, also demarcated the idyllic pastoral landscape of his childhood with his foster family in Gokul. At least five other Krishnalila patas, which may be attributed to the southwestern region of Medinipur, display a similar compositional strategy; pairing episodes of either side of the river, which flows downward. The variations offer useful comparison as the juxtapositions suggest an arena for innovation among painters. These include a scroll in the collection of the Gurusaday Museum (Acc. No. 1779); in the Los Angeles County Museum’s collection (M.71.1.32); The Victoria and Albert Museum (IS 109-1955); the Asutosh Museum of Calcutta University; and the private collection of David J. McCutchion. The Victoria and Albert Museum’s scroll was acquired by J. C. French, and bears a location of Tamluk, Medinipur Dist., where he likely took possession of the object during his travels.
 The earliest transcriptions of songs come from Gurusaday Datta, who travelled the region on work and spent his spare time indulging his curiosity about this artistic practice, among others. While we have photographic documentation of his interactions with such storytellers and some revealing intimate discussion, there are no archival records of such encounters. Here I have used Gurusaday Datta, Patua Sangit (Calcutta: Calcutta University Press, 1939); and David J. McCutchion and Suhrid Bhowmik, Patuas and Patua Art in Bengal (Calcutta: Firma KLM, 1999).
 Less often, the singers were commissioned to perform in the courtyards of elite residences as part of festival celebrations.
 The words of songs indicate that this could take the form of material goods such as rice or clothes.
 Among living practitioners, the resourcefulness in spontaneously customising a performance can include not only the selection of themes, but also choice of scrolls for display. A savvy practitioner astutely assesses the audience’s interest level, for example, to decide whether to pull out a longer or shorter version of the same story, or one with more detailed imagery versus a quick rendition; one with ‘natural’ or bolder colours, or what is often termed a more ‘tribal’ style. Gurupada Chitrakar of Noya, for example, explained that a more sophisticated urban elite clientele and foreigners prefer muted colours and an earthy palette (oral communication, Philadelphia, in 1997). They want natural dyes, which some patuas have begun to use instead of the commercial paints with a local resin or gum for binder. Such adaptation can be understood as a return to older practices, but can also amount to an invented tradition.
 As with most street musicians and entertainers, subsistence on voluntary contributions is precarious and requires ingenuity on the part of patuas; today the painters are quite upfront about discussing such issues with scholars. Beatrix Hauser has elaborated on the provisional nature of a performance in ‘From Oral Tradition to “Folk Art”: Reevaluating Bengali Scroll Paintings’, Asian Folklore Studies 61:1 (2002): pp.105–122.
 McCutchion and Bhowmik, Patuas and Patua Art, pp. 61–121.
 A second patua song elaborates on the craftiness of the charming Krishna, playing on word banka (not straight), to describe both the sinuous, curvaceous pose of the typical dancing Krishna icon and his crooked ways:At the base of the kadamba tree the urbane (naagar) Kanai (Krishna) dallies
Wild flowers gathered into garlands (vanamala) adorn his neck.
Arms crooked, legs twisted, torso twisted,
Twisted from the anklets of his feet to the crown of his topknot.
Loving maidens fix the topknot into wondrous creations
Entrancing the sixteen thousand alluring gopis of Braj. (Datta, Song 5, p. 17, my translation).
 Song of Triloktarini Chitrakar of Baliya, collected by Gurusaday Datta, pp.1–2. (My translation).
 Vaishnava belief and practice in Bengal conceptualises five emotional states for experiencing Krishna, graded in a hierarchy of increasing intensity and intimacy: santa, the peaceful condition in which the worshipper regards herself or himself as low and insignificant in relation to the supreme, omnipotent deity; dasya, the condition of devotion towards one’s master, sakhya, the state of friendship; vatsalya, a bond of caring affection between mother and child, and madhurya, the sweetness of love as Radha had offered him.
 The arresting figure of Vasudeva, with his head turned to the side, shares in a distinctive pictorial tradition associated with the Santal tribal community’s funerary practices. After a death, the jadupatua (magician-painter) delineated the pupil in the large single eye of similar seated figures, identifying the image with the dead person, and activating the spirit’s safe journey from the body of the deceased to its destination. Here that wide-eyed stare is employed to a different end; that of divine vision, albeit an equally perilous journey. The juxtaposition of figures in different scale and profile in this scene demonstrates the painter’s inventiveness in underscoring Vasudeva’s extraordinary experience. The scroll shares many other features with jadupatua pictorial style including the rendition of Devaki’s bed; the draping of saris; and the delineation of Vasudeva’s bare chest and diaphragm, and hair. Just how widespread this style, associated with Santhal pata subject matter, may have been at the turn of the twentieth century, remains to be studied carefully. On this tradition, see Amitabh Sengupta, Scroll Paintings of Bengal: Art in the Village (Bloomingdale, IN: Author House, 2012).
 The floating baby Krishna shares compositional elements with pictorial representations of Krishna on a leaf as encountered by the sage Markandeya. Examples include the Philadelphia Museum of Art folios, Markandeya discovers Krishna on the Peepal Leaf and The Vision of Markandeya (acc. no. 1955-11-1; 1994-148-492).
 While several scrolls insert this detail to the transporting of the baby, other visual genres such as terracotta imagery from previous centuries, or contemporary embroidered and printed versions of such episodes do not. The implication that this elaboration for dramatic effect is a patua interpolation by painters and song composers cannot be ruled out.
 By the time this scroll was created, these compositions had become easily recognisable, visualised on temples since the earliest temples dedicated to Krishna in the seventeenth century. For comparison, see for example, my discussion in ‘At Home in Bengal: Krishna’s Lila as Performing a Vernacular Bhagavata’, Journal of Hindu Studies 11.2 Special Issue New Directions in Bhagavata Purana Studies II, edited by Ravi Gupta and Kenneth Valpey. (August 2018): pp. 133–150. For a manuscript illustration, see, for example, the Isarda folio in the San Diego Museum of Art collection (1990.585), discussed most recently by Neeraja Poddar, ‘Incarnations of the Bhagavata’, in Marika Sardar (ed.) Epic Tales from Ancient India: Paintings from The San Diego Museum of Art (San Diego: San Diego Museum of Art), pp. 28–38, 40–63.
 Radha’s origins and her relationship to Krishna as revealed in Gaudiya and earlier Vaishnava texts has been discussed at great length by several scholars including Sasibhusan Dasgupta, Sriradhar Kramabikasa (Calcutta: E. Mukherji, 1953); Sushil Kumar De, Early History of the Vaisnava faith and movement in Bengal: from Sanskrit and Bengali sources (Calcutta: Firma KLM Private Limited; Columbia, MO: Distributed by South Asia Books, 1986); Edward Dimock, The Place of the Hidden Moon: Erotic Mysticism in the Vaiarlava-Sahajiya cult of Bengal (Chicago and London: University of Chicago Press, 1966); Daniel H. Ingalls and Milton Singer (eds), Krishna: Myths, Rites and Attitudes (Chicago and London: University of Chicago Press, 1985); J. A. B. Van Buitenen, ‘On the Archaism of the Bhāgavata Purāna’, in Krishna: Myths, Rites, and Attitudes, pp. 3–40; Charlotte Vaudeville, ‘Evolution of Love-Symbolism in Bhagavatism’, Journal of the American Oriental Society 82:1 (1962): pp. 31–40; Sumanta Banerjee, Appropriation of a Folk-Heroine: Radha in Mediaeval Bengali Vaishnavite Culture (Shimla: Indian Institute of Advanced Study, 1993); Norvin Hein, The Miracle Plays of Mathura (New Haven: Yale University Press, 1972), pp. 163–271; A. K. Majumdar, ‘A Note on the Development of the Radha Cult’, Annals of the Bhandarkar Institute 36 (1955): pp. 231–257; Barbara Stoler Miller, The Gitagovinda of Jayadeva: Love Song of the Dark Lord (New York: Columbia University Press, 1977; First Indian Edition Delhi: Motilal Banarsidass, 1984), pp. 26–37.
 The songs often narrate events that were not part of the Bhagavata Purana, but rather episodes elaborated by local poets such as the fourteenth-century luminaries, Chandidas and Vidyapati. Here I cite from the following translations: Baru Chandidas, Singing the Glory of Lord Krishna: The Srikrsnakirtana, (trans.) Miriam H. Klaiman (Chico, CA: Scholars Press, 1984); W. G. Archer (ed.), Love Songs of Vidyāpati (trans.) Deben Bhattacharya (New York: Grove Press, 1970).
 Song of Panchanan Chitrakar of Panuari, Datta, Patua Sangit, 22 (my translation).
 This artist has chosen to depart from more common renditions of the crawling baby between the tree trunks. Instead, Krishna is presented here among leafy branches at the top of the tree, much like he is located in another better-known episode of stealing the clothes of bathing women. Recycling motifs with minor variation to dramatic effect is not unusual. The song typically offers clarification. Repetition, moreover, serves to draw attention to similarities and differences. For example, a scroll dedicated to the Chandimangal narrative of the merchant Dhanapati and his son Srimanta (Philadelphia Museum of Art, 1969-131-1a, b) uses this strategy to visualise the same voyages undertaken by father and son, but to different outcomes due to their divergent relationships to the goddess. The differentiation of the identical compositions would have been elaborated in song.
 Song of Panchanan Chitrakar of Panuari, Datta, Patua Sangit, 22 (my translation).
 These songs give further nuance and play with the rich corpus of devotional poetry that continues to be sung among Krishna devotees in Bengal.
 That duality, at the heart of Vaishnava belief and practice, finds expression in songs that present Krishna asserting his divine stature in protest when she employs him as a porter. For example, in the Bharakhanda of Chandidas’ Srikrishnakirtan, Krishna says:
It’s not the load, but the shame, that’s oppressive.
How shall I live in society, Radha?
Don’t say it, Radha, don’t say it. What shame
It would be for Gadadhara, bearing a burden!
I killed the Cart-Monster, haven’t you heard it?
There were two arjuna trees I uprooted.
I have descended to massacre Kamsa.
Now shall I carry your burden of produce?
Peddling these milk goods has addled your judgment;
You’re not aware I am God in the highest.
You are extremely audacious, young cowmaid…
 Gurusaday Datta does not give us a name, but only that the song was collected from a Santhal patua, Patua Sangit, p. 40.
 On this episode, see Charlotte Vaudeville, ‘The Govardhan Myth in Northern India’, Indo-Iranian Journal 22:1 (1980): pp. 1–45; John Stratton Hawley, ‘Krishna’s Cosmic Victories’, Journal of the American Academy of Religions 47:2 (1979): pp. 201–221. In consonance with the animation of the mountain in ritual, another patua song that explicates the episode at greater length also makes no mention of Indra, who was defeated and banished, and his worship stopped:
Saying kill, kill, the mountain began to crumble and fall
Twelve cowherd boys held up the mountain by the touch of a finger
From then the Lord was named Giridhari [lifter of the mountain].
 David L. Haberman discusses such practices in ‘Vraja: A Place in the Heart’, in Re-discovering Braj: International Association of the Vrindaban Research Institute Bulletin 14 (1988), pp. 19–25 and Acting as a Way of Salvation: A Study of Raganuga Bhakti Sadhana. (New York: Oxford University Press, 1988).
 Song of Panchanan Chitrakar of Panuari, Datta, Patua Sangit, 22 (my translation).
 The Maithili poet Vidyapati, for example, hints at Radha’s complicit participation in the signature line (bhanita). See Vidyapati, Love Songs of Vidyāpati, 133.
 Song of Kirti Chitrakar of Dumka, Datta, Patua Sangit, 37 (my translation).
 McCutchion and Bhowmik, Patuas and Patua Art in Bengal, 63.
 Brahma realises that Krishna had not only figured out what he had done, but used his divine power to replicate the cows and cowherd boys, so that when he returned he was bemused to find the cows and cowherd exactly where he had initially encountered them.
 McCutchion and Bhowmik, Patuas and Patua Art in Bengal, 64.
 The scroll ends abruptly here, but clearly there were additional scenes that followed as the frayed edges indicate. Correspondingly, the complete scrolls, such as those of the Victoria and Albert Museum and Gurusaday Museum, offer possibilities to imaginatively extend the panels to completion. Other typical choices include: the image of Yama, god of death; the tortures in hell; an array of deities present to acknowledge the superiority of the god to whom the scroll is devoted; or another iconic image of the deity.