Uniting Bharata's kinetic grammar with its stone testimony: Chidambaram, Tanjavur, Kumbakonam, Tiruvannamalai, Vṛddhācalam.
करणानां शिलालेखीय एवं कलेतिहास-विवरणम्A karaṇa is not a metaphor for stone; it is a set of coordinates — a standing base (sthāna), a leg action (cārī), and a pair of arm actions (nṛtta-hasta) — precise enough that a sculptor working in granite eight centuries ago and a dancer training today are, when both are faithful to the source, describing the same instant of motion from two different materials. This register exists to hold those two descriptions side by side: the śloka that defines the instant, and the stone that freezes it. What follows does not treat the temple as illustration of the text, nor the text as caption for the temple. Each is read as an independent witness, and the register's task is to show where the two witnesses agree, where they diverge, and what that agreement or divergence tells us about how the tradition itself understood its own inheritance.
The Nāṭyaśāstra locates the karaṇa system inside a specific transmission narrative rather than presenting it as an anonymous technical inventory. In the account woven through the text's own frame story, Śiva performs the tāṇḍava — a mode of vigorous, cosmic dance distinct from Pārvatī's lāsya — and it is Tāṇḍu, a sage attached to Śiva's retinue, who is instructed to carry this vocabulary of movement to Bharata, the sage-dramaturge credited with compiling the Nāṭyaśāstra itself. The fourth chapter, where the 108 karaṇas are enumerated, is accordingly titled the Tāṇḍava Lakṣaṇa — "the characteristics of tāṇḍava" — and opens not as a dance manual but as an act of homage, describing Śiva's own performance of the karaṇa sequence before it is handed down as instruction. This origin narrative does real interpretive work: it explains why the karaṇa catalogue is treated, uniquely among the Nāṭyaśāstra's technical inventories, as a closed and non-negotiable set. A grammar of hand-gestures (hasta) or facial expressions (mukhaja) can be regionally elaborated because it encodes communicable meaning subject to audience convention; a karaṇa, understood as a fragment of Śiva's own cosmic motion, cannot be elaborated in the same way without claiming to improve upon the god's own dance. The mythological frame therefore functions as a textual mechanism of closure — it is the reason 108 remains 108, rather than a number that later commentators felt free to extend.
It is this same narrative that licenses the architectural program discussed in Part Three: a temple gopuram or vimāna carved with the 108 karaṇas is not merely displaying a dance manual in stone. It is depicting Śiva Naṭarāja performing the tāṇḍava at the moment of its transmission to Tāṇḍu and Bharata — which is why, at sites such as Chidambaram, the karaṇa panels are positioned as an ornamental program surrounding or approaching the sanctum of Naṭarāja himself, and not as a freestanding didactic frieze unconnected to the temple's central deity.
Bharata's own working definition, restated by every subsequent commentator on the chapter, treats the karaṇa as a composite unit built from exactly three simultaneous components:
Hasta (the arm/hand configuration, more precisely the nṛtta-hasta or "dance-hand," distinguished from the semantically loaded hastas of abhinaya) + Pāda (the leg action, technically the cārī) + Sthāna (the standing posture or base) = Karaṇa. — restated from Nāṭyaśāstra, Ch. 4, Tāṇḍava Lakṣaṇa, opening definitional verses
The formula looks additive, but it is not: a karaṇa is not three actions performed in sequence but one coordinated event in which the sthāna, cārī, and pair of nṛtta-hastas occur together, within a single count of time, as one indivisible unit of movement. This is the feature that makes the karaṇa the correct minimal unit for the epigraphic register: a sculptor cannot carve "the leg action" and "the arm action" as two separate reliefs and expect a viewer to reconstruct the karaṇa by mental addition. A single panel must therefore capture all three components locked into the same instant — which is precisely the discipline the Chola- and Nayaka-period sculptors at Chidambaram, Tanjavur, and Kumbakonam impose on themselves, and precisely the criterion by which a panel can be identified, or misidentified, as a specific numbered karaṇa.
Two further technical terms belong to this anatomy and recur throughout Parts Two through Four. The cārī itself is a compound unit — a specific foot position (of which the Nāṭyaśāstra recognises sixteen) joined to a specific leg-raising or leg-extending action — so that the "pāda" element of the karaṇa formula is already, at one remove, a small combinatorial system in its own right. The nṛtta-hasta, meanwhile, is explicitly distinguished in the commentarial tradition from the śākhā and nṛtya hastas used in abhinaya: a nṛtta-hasta carries no representational or narrative content and exists purely as a formal, ornamental counterpart to the leg and torso action, which is the reason a karaṇa can be correctly executed, and correctly carved, without any accompanying facial abhinaya at all.
The distinction between mārgi and deśī — pan-Indian, texturally sanctioned classical practice against regionally inflected, vernacular practice — is not original to dance criticism; it is inherited from a broader classical vocabulary distinguishing the sanctioned "path" (mārga) of textual authority from the "country-specific" (deśa) elaborations local courts and communities generated around it. Applied to the karaṇa system specifically, the distinction carries unusual weight, because it is the karaṇas, more than any other layer of Bharata's technical vocabulary, that the tradition treats as belonging wholly to mārgi practice. A regional court dance, a temple's own local liturgical performance tradition, or a folk form is free — and is documented across South Indian performance history as having exercised that freedom liberally — to develop its own deśī vocabulary of gesture, gait, and gait-ornament. The 108 karaṇas, by contrast, are presented and epigraphically reproduced as a single pan-Indian standard, identical in principle whether the temple stands in Chidambaram, Tanjavur, or (for comparative purposes discussed in Part Three) in temples well outside Tamil-speaking country.
This is precisely what makes the sculptural program at these five Tamil temples art-historically remarkable: each site is a deśī institution — a specific royal or local devotional foundation, funded by a specific dynasty, embedded in a specific regional liturgical calendar — choosing to reproduce a mārgi textual standard in permanent stone. The panels are, in this sense, a deliberate act of claiming pan-Indian textual authority for a regional royal monument, and the variations in execution across the five sites (examined panel by panel in Part Three) can be read as the residue of deśī sculptural workshops each interpreting a mārgi textual mandate through their own regional idiom, tolerances, and stylistic conventions.
Nṛtta — pure, non-representational dance, as distinct from nṛtya (expressional dance carrying rasa and bhāva) and nāṭya (dramatic dance embedded in plot) — is the category within which the karaṇa system operates. A karaṇa carries no independent narrative meaning; its aesthetic function is formal and rhythmic rather than referential, and this is precisely why it can be studied, and carved, as a closed catalogue rather than as an open-ended expressive vocabulary. Above the level of the individual karaṇa sits the aṅgahāra: a fixed sequence of karaṇas, of which Bharata enumerates thirty-two, each aṅgahāra combining a specific and named set of karaṇas in a specific and non-negotiable order. Sequence-order is not a stylistic preference at this level; a performer who executes the correct karaṇas belonging to a given aṅgahāra but in the wrong order has, by the tradition's own reckoning, performed a different composition entirely, not a looser version of the intended one.
Above the aṅgahāra sits the maṇḍala, or more precisely the recognisable higher-order groupings and, for ensemble work, the piṇḍībandha formations in which several performers coordinate their individually executed aṅgahāra sequences into a single spatially organised composition. This three-tier hierarchy — karaṇa, aṅgahāra, piṇḍībandha — is the compositional architecture the epigraphic record must be read against: a single gopuram panel typically depicts one karaṇa in isolation, but the ordering of panels along a gopuram's tiers, discussed at each site in Part Three, is frequently not arbitrary, and where the temple's own panel sequence tracks the Nāṭyaśāstra's own textual sequence of the 108, the sculptural program is implicitly encoding aṅgahāra-level sequence information in the architecture itself, not merely a gallery of isolated postures.
Beneath the karaṇa lies a further, smaller layer of movement vocabulary that the epigraphic record must resolve at an even finer grain: the recakas, four distinct rotational or circling actions assigned respectively to the hand (hasta-recaka), foot (pāda-recaka), waist (kaṭi-recaka), and neck (grīvā-recaka). A recaka is not itself a karaṇa; it is a distinguishing rotational quality that a karaṇa's own sthāna, cārī, or nṛtta-hasta component may or may not incorporate, and it is frequently the recaka alone that differentiates two otherwise similar karaṇas in the catalogue — a distinction that places unusually demanding requirements on a sculptor, who must convey a rotational, time-based quality (a limb captured mid-turn, rather than at rest in a turned position) using a medium with no literal capacity to depict motion directly.
Bhramarī — spinning or turning movement, executed as a rotation of the whole body about its own vertical axis — belongs to a related but distinct category, associated with several of the more advanced karaṇas in the later portions of the catalogue (examined in Part Four, §24) and posing the same representational problem in an intensified form: a full-body spin is, by definition, a movement no single frozen instant can fully convey, which is one reason sculptors at several of the sites discussed below appear to select, for bhramarī-type karaṇas specifically, the instant of maximum extension or maximum torsion within the turn, rather than an arbitrary point in its rotation — an interpretive choice examined directly in §18's treatment of "frozen motion." Maṇḍala movements, finally, refer to the larger spatial patterns — diagonal, circular, and cruciform traversals of the performance space — that several aṅgahāras trace across the stage as a whole; these are visible in the epigraphic record only indirectly, through the sequential arrangement of karaṇa panels along a gopuram's own spatial program, discussed site by site in Part Three.
Chapter Four occupies a structurally unusual position within the Nāṭyaśāstra as a whole. Where the surrounding chapters build a general theory of dramatic production — rasa and bhāva in Chapter Six and Seven, the hasta-mudrā system in Chapter Nine, the cārīs and maṇḍalas in Chapters Ten through Twelve — Chapter Four is narrower and more technical than its neighbours: it is, in effect, a closed appendix of exactly 108 named definitions, each given in compact śloka form, typically specifying the sthāna, cārī, and hasta components in a single or double verse with minimal elaboration and no accompanying dramaturgical commentary within the root text itself. This compactness is precisely what necessitated the commentarial tradition examined in Part Two: a verse that names a sthāna and a cārī in six or eight words, without illustration, is legible to a reader who already possesses the physical vocabulary those technical terms presuppose, and largely illegible to one who does not — which is exactly the gap Abhinavagupta's commentary, and later the temple sculptural program itself, exists to close.
The chapter's placement early in the text, before the more expansive treatment of abhinaya in later chapters, also carries an implicit theoretical claim: that pure kinetic form (nṛtta) is treated as logically and pedagogically prior to expressional dance (nṛtya) and dramatic dance (nāṭya), a sequencing later inherited by the training progression standard in classical dance pedagogy, in which a student's early years are given over substantially to adavu-level and karaṇa-adjacent pure movement training before expressional repertoire is introduced.
The karaṇa panels at the sites examined in Part Three are not, in the overwhelming majority of cases, accompanied by narrative inscriptions describing the dance program itself; the panels function as visual rather than textual documents. Where inscriptional material relevant to the karaṇa program does survive at these temples, it belongs to a different and more general epigraphic category: donative, dedicatory, and administrative inscriptions recording temple grants, dancer endowments (the devadāsī institution and its associated land and salary grants), and dynastic patronage, typically rendered in Grantha script — the script developed specifically for writing Sanskrit in Tamil-speaking regions, distinct from the Tamil script used for vernacular administrative and literary inscriptions at the same sites, and often appearing alongside Tamil-script inscriptions on the same temple walls, addressing different registers of content.
This bilingual and biscriptural epigraphic environment matters directly to the karaṇa program's own interpretation: Grantha-script Sanskrit inscriptions at several of these temples record the formal dedication of dance performance to the temple deity, the endowment of dancers attached to specific ritual occasions, and — at Chidambaram specifically — inscriptional references to ritual tāṇḍava performance connected to the temple's own liturgical calendar. These inscriptions do not gloss individual karaṇa panels by name, but they establish, independently of the sculptural evidence itself, that karaṇa-vocabulary dance was an actively maintained ritual practice at these sites during the same centuries the sculptural programs were commissioned — evidence that the 108-karaṇa reliefs are not antiquarian décor referencing a defunct textual tradition, but the architectural counterpart of dance that was, at the time of carving, still being performed in the temple's own precincts.
The single most important interpretive bridge between Bharata's compressed fourth-chapter ślokas and any concrete, reconstructible physical movement is Abhinavagupta's Abhinavabhāratī, the tenth- to eleventh-century commentary on the Nāṭyaśāstra composed in Kashmir — geographically and historically distant from the Tamil temple sites this register is chiefly concerned with, and this distance is itself methodologically significant: it means the commentarial tradition that later South Indian sculptors and dance masters drew upon to interpret the karaṇa ślokas was, by the time it reached Tamil country, already a text that had travelled across a substantial span of India and several centuries, carrying with it Kashmir Śaiva interpretive habits that shaped how individual karaṇa verses were read.
Abhinavagupta's method with the karaṇa chapter is characteristically exegetical rather than purely lexical: where Bharata's root verse names a sthāna and cārī in minimal technical shorthand, Abhinavagupta typically supplies the fuller physical elaboration a practising reader would need — clarifying, for instance, which of the sixteen cārīs a compressed verse-reference is invoking, resolving ambiguous compound terms that could denote more than one recognised sthāna, and, in a number of documented cases, correcting or flagging manuscript variants where the root text's transmission had become corrupt. It is precisely because Abhinavagupta's commentary supplies this physical specificity that modern reconstruction projects (examined in §12) treat the Abhinavabhāratī as an indispensable second witness alongside the root text itself — the karaṇa ślokas alone are, in a significant number of cases, underdetermined as physical instructions without the commentarial gloss resolving which specific cārī or sthāna a given technical term denotes.
The compositional register of Chapter Four is closer to sūtra literature than to the more expansive descriptive verse found elsewhere in the Nāṭyaśāstra: each karaṇa definition is compressed into the smallest verse unit capable of naming its three components, a style that trades descriptive completeness for mnemonic and recitational efficiency. This sūtra-like compression served a specific transmission function in an oral-textual culture where the fourth chapter needed to be memorised and recited by a guru transmitting the system to a student's body rather than merely read silently as reference material — a compact couplet is retrievable from memory in a way a fuller descriptive paragraph is not, and the terseness of the verse is best understood as designed for a pedagogical situation in which the verse functions as a retrieval cue for knowledge already substantially held in the trained body, rather than as a self-sufficient written instruction for a reader with no prior physical training.
This has a direct consequence for how the epigraphic evidence should be weighted against the textual evidence: because the root verses are compressed to the point of relying on an already-trained reader, a sculptural panel carved by an artisan working alongside a dance master — which the epigraphic and inscriptional record, per §7, suggests was the likely production context at several of these temples — may in some cases preserve physical detail the terse root verse alone does not fully specify, making the stone not merely an illustration of the text but, in specific instances, a supplementary witness to physical practice that fills gaps the sūtra-style verse leaves open.
Two later texts function as essential comparative controls against Bharata's own fourth chapter. Nandikeśvara's Abhinaya Darpaṇa (probably compiled well after the Nāṭyaśāstra, though drawing on the same broad textual family) restates a substantial portion of the karaṇa-adjacent vocabulary in a more accessible, less compressed idiom, and is documented as having served — considerably more directly than Bharata's own text — as the working manual actually consulted by later generations of practising dance masters, including within the Tamil devadāsī and naṭṭuvanār teaching lineages associated with several of the temples examined in Part Three. Where the Abhinaya Darpaṇa's own treatment of related movement categories diverges from Bharata's fourth chapter — as it does at several specific points, particularly in nomenclature and in the ordering of certain cārīs — the divergence is itself evidence of a living, regionally adapted transmission rather than a simple error in one text or the other, and comparative readings across the two sources are frequently what allow a modern reconstructor to resolve an otherwise ambiguous root verse.
Śārṅgadeva's thirteenth-century Saṅgīta Ratnākara, composed in the Deccan under Yādava patronage, supplies a different and equally necessary control: it is the major post-Bharata text to treat dance systematically alongside music and, crucially, alongside a developed theory of tāla, allowing a comparative reader to check whether Bharata's own, considerably less explicit fourth-chapter treatment of rhythmic placement is consistent with the more elaborated rhythmic theory Śārṅgadeva later supplies. Because the Saṅgīta Ratnākara post-dates Bharata by many centuries and reflects its own Deccan-regional synthesis, it cannot be read as a transparent gloss on the Nāṭyaśāstra's original intent, but it remains indispensable as evidence for how the karaṇa-tāla relationship was understood in precisely the historical window — the centuries immediately preceding the major Tamil temple building programs discussed in Part Three — during which the sculptural projects examined below were conceived and executed.
Structurally, the 108 karaṇas of Chapter Four are not presented in the root text as a randomly ordered list; the sequence in which Bharata names them tracks, in significant part, the sequence in which they are subsequently deployed across the thirty-two aṅgahāras enumerated later in the same chapter, so that the catalogue functions simultaneously as a definitional glossary and as an implicit index into the compositional units built from it. This has a direct methodological consequence for reading the temple sculptural programs examined in Part Three: where a gopuram or vimāna preserves its karaṇa panels in an order matching, or closely approximating, the Nāṭyaśāstra's own textual sequence — as is documented, with some local variation, at Chidambaram and, in modified form, at Tiruvannamalai — the architectural program is best read as a direct spatial transposition of the textual catalogue, tier by tier up the gopuram's own vertical structure standing in for verse by verse down the manuscript page.
Where a site's panel sequence instead departs from the textual order — as documented, with specific local logic, at Kumbakonam and in the incomplete program at Vṛddhācalam (examined in §15 and §17) — the departure is itself diagnostic evidence about the sculptural program's own compositional priorities: iconographic pairing with adjacent deity or narrative panels, structural constraints of the gopuram's own tier divisions, or, in the Vṛddhācalam case, an incomplete commission rather than a deliberate reordering. Mapping each site's actual panel sequence against Bharata's textual sequence is accordingly the single most consequential piece of comparative groundwork underlying every site-specific identification offered in Part Three.
The single most consequential modern contribution to reading the karaṇa chapter against its sculptural evidence is the doctoral and subsequent decades-long research program of Dr. Padma Subrahmanyam, whose method — formalised in her doctoral thesis and refined across her subsequent Karaṇas in Indian Dance Sculpture and related publications — established a specific, replicable procedure for identifying a given temple panel with a specific numbered karaṇa in Bharata's list. The method proceeds by triangulation rather than by visual resemblance alone: a candidate panel is checked against (a) the root Nāṭyaśāstra verse's named sthāna and cārī components, (b) Abhinavagupta's commentarial gloss resolving any ambiguous terminology in that verse, and (c) the panel's own physical configuration of base, torso, and limb positions, with a positive identification withheld unless the sculptural evidence and the textual-commentarial evidence converge on the same named karaṇa rather than merely being compatible with it.
This triangulated method is what allowed Subrahmanyam to propose systematic, panel-by-panel identifications across the Chidambaram gopuram program (§13) and, significantly, to recognise the long-overlooked Brihadīśvara vimāna reliefs (§14) as a second, independent, and near-complete 108-karaṇa cycle — a discovery that transformed the art-historical status of the karaṇa sculptural program from a single celebrated instance at Chidambaram into a comparative corpus spanning multiple Chola and Chola-successor sites. A further, complementary strand of her scholarship involved the practical reconstruction and re-performance of individual karaṇas identified through this method within her own Bharatanatyam and Bhāratanṛtyam practice, treating the sculptural evidence not merely as an art-historical artefact to be catalogued but as a physically re-enactable record — a methodological position directly continuous with §25's discussion of the karaṇas' afterlife in contemporary classical performance.
The Chidambaram Naṭarāja temple carries the most celebrated and most textually complete of the five sculptural programs examined in this register. The 108 karaṇas are carved as a systematic series in the Naṭarāja Sabhā's own eastern and western gopurams — most extensively catalogued on the eastern (Cit Sabhā-facing) gopuram — where each karaṇa occupies its own individual niche or panel, typically accompanied by an identifying inscriptional label in Grantha script naming the karaṇa depicted, a feature that distinguishes Chidambaram from every other site in this register and is the single reason Chidambaram functions as the anchor site against which all other identifications are calibrated. Where a panel's own physical configuration is ambiguous or worn, the accompanying inscriptional label at Chidambaram frequently resolves the identification directly, without requiring the fuller triangulated method §12 describes for unlabelled sites.
The theological logic of the program's placement is direct rather than incidental: Chidambaram is, in Śaiva theology, the site of Śiva's own Naṭarāja manifestation and of the Ānanda Tāṇḍava specifically, so that a gopuram program depicting all 108 karaṇas of the Tāṇḍava Lakṣaṇa chapter functions as an architectural restatement, at the temple's own threshold, of the mythological transmission narrative examined in §1 — a visitor approaching the sanctum of the dancing Śiva passes, quite literally, through a gateway inscribed with the complete grammar of his dance. The panels are consistently read by Subrahmanyam's reconstruction as following the Nāṭyaśāstra's own textual sequence with high fidelity, making Chidambaram the primary evidentiary base for the sequence-mapping method described in §11.
The Brihadīśvara temple, commissioned under Rājarāja I Chola and completed in the early eleventh century, preserves a second, near-complete 108-karaṇa cycle carved on the temple's own vimāna — the towering pyramidal superstructure over the main sanctum — in a location considerably less accessible to casual observation than Chidambaram's gopuram panels, positioned high on the vimāna's upper tiers rather than at ground-approach level. This placement is directly responsible for the relief cycle's long art-historical obscurity: the panels were not systematically identified, catalogued, and connected to the Nāṭyaśāstra's own karaṇa sequence until Padma Subrahmanyam's twentieth-century research program undertook the sustained close-range documentation the vimāna's height had previously discouraged.
The Tanjavur cycle's significance is comparative rather than merely additive: because it was carved under direct Chola imperial patronage at a securely dated site, contemporaneous with or shortly preceding the main phase of Chidambaram's own gopuram construction, it supplies independent confirmation that the systematic 108-karaṇa sculptural program was an established Chola court convention rather than a single unrepeated undertaking at Chidambaram alone. Discrepancies between the Tanjavur panel sequence and the Chidambaram sequence — documented at several specific points in Subrahmanyam's comparative analysis — are correspondingly significant evidence for how much interpretive latitude a Chola-period sculptural workshop retained even while working from what was, in both cases, understood as the same authoritative textual source.
The Sarangapani temple's karaṇa program, carved on the temple's external facade rather than on a gopuram or vimāna proper, represents a Vaiṣṇava rather than Śaiva sculptural context for the same iconographic vocabulary — a significant art-historical fact in its own right, since it demonstrates that the 108-karaṇa program had, by the period of the facade's construction, become sufficiently detached from its originating Śiva-tāṇḍava mythological frame (§1) to be adopted as a general ornamental and technical program within a Vaiṣṇava temple's own decorative scheme, without requiring a Naṭarāja-specific theological justification of the kind that anchors the Chidambaram program so directly.
The facade placement itself produces a distinct set of interpretive conditions from the gopuram-tier placement at Chidambaram and Tiruvannamalai: facade reliefs are typically arranged in a single or double horizontal register at a consistent viewing height, rather than distributed across multiple ascending gopuram tiers, which affects both how a devotee historically encountered the program (as a continuous horizontal narrative band, walkable and legible in a single circumambulatory pass, rather than as a vertically ascending sequence requiring the visitor to look progressively upward) and how confidently a modern researcher can establish panel-to-panel sequence, since a single horizontal register is considerably less ambiguous, tier-boundary and orientation-wise, than a multi-tier gopuram program.
The Arunachaleśvara temple's karaṇa program, carved into the temple's Rāja Gopuram, presents the register's clearest instance of a systematic, tier-by-tier architectural arrangement of the full sequence: the panels are distributed methodically across the gopuram's ascending storeys in a manner that Subrahmanyam's and subsequent researchers' documentation reads as a deliberate structural device, using the gopuram's own multi-tiered vertical architecture as a physical index for the aṅgahāra-level sequencing discussed in §11 — a devotee's ascent of visual attention up the gopuram's tiers correlating, tier by tier, with progression through the textual catalogue's own sequence.
This systematic arrangement makes Tiruvannamalai a particularly valuable comparative check against Chidambaram's own sequence-fidelity: because the Rāja Gopuram's tier divisions impose their own architectural logic (a fixed, finite number of panels per tier, dictated by the tier's own physical dimensions rather than by the textual catalogue's own natural divisions), any point at which the Tiruvannamalai sequence must depart from strict textual order to accommodate the architecture's own tier boundaries is directly observable, and supplies concrete evidence for how a Vijayanagara-period (the Rāja Gopuram's approximate construction era) sculptural workshop negotiated the tension between architectural constraint and textual fidelity — a negotiation the Chidambaram program's own gopuram structure handled somewhat differently, given its different tier proportions and earlier construction period.
The Vṛddhagiriśvara temple at Vṛddhācalam preserves what is, in the current state of scholarship, the register's most significant negative case: a karaṇa sculptural program that stops at 101 panels rather than completing the full 108, presenting a documented and unresolved gap in the otherwise systematic sequence. Two explanations are current in the art-historical literature, and this register does not adjudicate definitively between them: the incompletion may reflect an interrupted or abandoned commission — a not-uncommon fate for large-scale South Indian temple sculptural programs dependent on continuous royal or local patronage across what could be a multi-generational construction timeline — or it may reflect a deliberate, locally specific decision to omit certain karaṇas, possibly on ritual or iconographic grounds specific to this temple's own devotional context, a possibility that remains, at present, unconfirmed by any surviving inscriptional or textual explanation at the site itself.
What can be established with more confidence is the diagnostic value of the gap itself: because the surviving 101 panels can be checked against Chidambaram's fuller and inscriptionally labelled sequence, the specific identity of the missing seven karaṇas is recoverable by comparative elimination, and the pattern of which karaṇas are absent — rather than a uniform truncation of the sequence's final entries, the omissions are distributed at several points through the catalogue — is itself evidence weighing against a simple "construction stopped partway through" explanation and lending some support to the possibility of deliberate, selective omission, though this register presents that reading as the more probable rather than the conclusively established account.
Every panel examined across §13 through §17 confronts the same underlying representational problem: a karaṇa, by Bharata's own definition (§2), is a coordinated, momentary event, while a stone relief is, necessarily, a single static instant with no capacity for literal temporal extension. The sculptural tradition's solution to this problem is not uniform across the five sites, and the variation is itself art-historically informative. The dominant convention, documented most consistently at Chidambaram, selects the instant of maximum extension or maximum torsional tension within a given karaṇa's own trajectory — the point at which the sthāna, cārī, and nṛtta-hasta components are most clearly and unambiguously differentiated from a neutral standing position — reasoning, implicitly, that this is the instant from which a viewer already familiar with the movement's trajectory can most reliably reconstruct the whole action, both what preceded it and what follows.
A secondary and less common convention, visible in a smaller number of panels particularly among the more vigorous bhramarī-type and vṛścika (scorpion) karaṇas discussed further in §24, attempts a more explicit suggestion of motion itself — through asymmetric drapery treatment implying momentum, through slightly exaggerated torsional distortion of the torso beyond what any single static pose could naturally hold, or through the juxtaposition of two adjacent panels depicting sequential moments of what the Nāṭyaśāstra's own aṅgahāra sequencing (§4, §11) indicates were performed as a continuous pair — a device that effectively borrows filmic sequential-frame logic centuries before the concept existed under that name, and represents the sculptural tradition's most sophisticated solution to the frozen-motion problem this section identifies.
Although the 108-karaṇa catalogue is textually and mythologically anchored to Śiva's tāṇḍava (§1), the sculptural programs examined in this register are not, with the significant exception of certain Chidambaram panels depicting the central Naṭarāja icon itself, individually labelled as depictions of Śiva performing each specific karaṇa; the majority of panels across all five sites depict an unidentified or generically rendered male dancer figure, executing the karaṇa in question without the specific iconographic attributes (the flaming halo, the damaru, the encircled Apasmāra) that would mark the figure unambiguously as Naṭarāja rather than as a human performer or a gaṇa attendant demonstrating the movement.
This iconographic restraint is itself theologically and art-historically significant: it suggests the temples' own sculptural workshops distinguished consistently between the singular, theologically central Naṭarāja icon — reserved for specific, doctrinally load-bearing images such as the bronze mūla-mūrti and its principal stone representations — and the generic didactic figure appropriate to a technical catalogue illustrating the grammar of dance rather than depicting a specific theophany at every one of its 108 instances. Where Pārvatī appears within or adjacent to these programs, her presence is typically confined to specific narrative or paired panels rather than distributed evenly across the karaṇa sequence, consistent with the karaṇa catalogue's own textual attribution to Śiva's tāṇḍava specifically rather than to Pārvatī's lāsya, and confirming that the sculptural programs, across all five sites, were understood by their own commissioning workshops as illustrating a specific, gendered mythological narrative rather than a generic, ungendered dance-technical inventory.
The doctrinal, mythic, and definitional ground on which the epigraphic and sculptural record of Part Three, and the catalogue of Part Four, are read.
करणानां तात्त्विक-पौराणिक-मूलाधारःBefore a single temple panel can be read and before a single karaṇa can be catalogued, the system that governs both requires its own theoretical account: what a karaṇa is, where the tradition says it came from, what authority closes its number at exactly 108, and what larger compositional grammar it belongs to. This Part establishes that ground in full, so that Parts Two through Four can proceed as applied readings of a system already defined here with precision, rather than re-deriving first principles at every site and every entry.
Classical Śaiva mythography organises Śiva's dance activity into two structurally opposed modes rather than treating "dance" as an undifferentiated category. Tāṇḍava is characterised across the Purāṇic and āgamic literature as vigorous, forceful, and cosmologically consequential — the mode in which Śiva enacts creation, preservation, destruction, concealment, and grace, the five acts (pañcakṛtya) most fully theorised in the specifically Chidambaram-linked Śaiva Siddhānta tradition examined further in 1.4 below. Lāsya, by contrast, is characterised as gentler, more lyrical, and is associated in the mythological record with Pārvatī. The 108 karaṇas of the Nāṭyaśāstra's fourth chapter belong, both by the chapter's own title (Tāṇḍava Lakṣaṇa, "the characteristics of tāṇḍava") and by the narrative frame surrounding it, specifically to the tāṇḍava mode. This is not a neutral classificatory detail: it determines which iconographic register the sculptural programs of Part Three are entitled to invoke (a vigorous, masculine-coded dance vocabulary rather than Pārvatī's own lāsya repertoire), and it is the reason no comparably systematic 108-unit catalogue of lāsya movement survives attached to the same textual authority — the lāsya tradition is transmitted, historically, through a more diffuse and less codified body of regional practice rather than through a single closed Nāṭyaśāstra chapter.
The Nāṭyaśāstra's own frame narrative does not present Bharata as receiving the karaṇa vocabulary directly from Śiva; it interposes a named intermediary, the sage Tāṇḍu, whose own name is given by the tradition as the etymological source of "tāṇḍava" itself. Tāṇḍu is described as having been taught the dance directly by Śiva and subsequently instructed to transmit it to Bharata for inclusion in the Nāṭyaśāstra — a two-step chain (Śiva to Tāṇḍu, Tāṇḍu to Bharata) rather than a single divine dictation. This detail matters for how the text's own authority claim should be read: the karaṇas are presented not as Bharata's own compositional invention, nor as an anonymous accretion of practice codified after the fact, but as a specific act of textual transmission with a named intermediate link, structurally comparable to the way other Sanskrit śāstric traditions ground technical authority in a named teacher-lineage (paramparā) reaching back to a divine or semi-divine source, rather than in the compiling author's own independent expertise.
A recurring interpretive question this register must resolve before Part Four's catalogue can be approached responsibly is why the number of karaṇas is fixed at 108 and treated, across every commentarial and sculptural source examined in Parts Two and Three, as a closed rather than an extensible set — in marked contrast to other technical inventories within the same Nāṭyaśāstra (the hasta-mudrā system, for instance, is documented as having undergone regional elaboration well beyond the root text's own count without controversy). The mythological frame supplies the most direct available explanation: because the karaṇas are presented as a specific, complete performance already executed by Śiva himself and transmitted, rather than as an open descriptive category available for a human compiler's own extension, adding a 109th karaṇa would carry the theological implication of correcting or supplementing the god's own dance — a move the tradition's own internal logic forecloses in a way it does not foreclose for more clearly human-authored technical categories elsewhere in the text. The number 108 itself carries independent numerological weight within Śaiva and broader Indic tradition (the count of beads on a standard mālā, the number of Upaniṣads recognised in some enumerations, recurring temple-architectural counts), and this register treats the karaṇa count as continuous with, rather than incidental to, that broader numerological convention, without claiming a single, textually stated explanation for why 108 specifically was the number selected.
The Chidambaram tradition specifically elaborates Śiva's dance into the five cosmic acts referenced in 1.1 — sṛṣṭi (creation), sthiti (preservation), saṃhāra (destruction), tirobhāva (concealment), and anugraha (grace) — a theological schema most fully developed in the temple's own liturgical and philosophical tradition and closely associated with the Cidambara Māhātmya and related Śaiva Siddhānta commentary literature specific to the site. The karaṇa catalogue is not identical with this fivefold schema — the 108 karaṇas are units of pure kinetic form (nṛtta), while the pañcakṛtya is a cosmological interpretation of what that form, taken as a whole, accomplishes — but the two are presented by the temple's own theological tradition as continuous: the Ānanda Tāṇḍava performed at Chidambaram's own sanctum is read as simultaneously the complete kinetic event the 108 karaṇas catalogue and the complete cosmic act the pañcakṛtya schema interprets. This dual reading is precisely why the Chidambaram gopuram program (examined at length in Part Three) can function simultaneously as a technical dance-glossary and as a theological statement: the same stone sequence carries both meanings without contradiction, because the tradition itself does not treat the kinetic and cosmological registers as separate.
The mythological transmission narrative established in 1.1 through 1.4 does not remain confined to textual and liturgical tradition; it is the direct interpretive warrant for reading a gopuram's 108-karaṇa relief program as more than didactic illustration. A visitor's physical approach toward the Chidambaram sanctum, moving through a gopuram inscribed with the complete karaṇa sequence, is architecturally structured to re-enact, at the level of devotional experience, the same transmission the mythological narrative describes at the level of cosmic history — approaching the dancing god by first passing through the complete grammar of his dance. This is the theoretical basis on which Part Three's site-by-site treatment proceeds: the placement, ordering, and theological framing of each sculptural program is read not as a separate, secondary question from the mythological material established here, but as its direct architectural continuation.
A karaṇa is constituted by the simultaneous union of a sthāna (standing base), a cārī (leg action), and a pair of nṛtta-hastas (formal, non-narrative arm positions) — not their sequential combination, but their coincidence within a single indivisible instant of performed time. — restated from Nāṭyaśāstra Ch. 4, Tāṇḍava Lakṣaṇa, opening definitional matter, as consistently glossed by Abhinavagupta
The formula's apparent simplicity — three named components joined by addition — conceals a stricter requirement than ordinary language suggests. Bharata's own commentators are consistent in reading the "+" of the formula as denoting simultaneity of execution rather than sequence: a performer who executes a correct sthāna, then a correct cārī, then a correct nṛtta-hasta pair, one after another, has not performed a karaṇa at all, however accurately each individual component is rendered, because the definitional requirement is that all three occur as one coordinated event within a single count. This is the single most consequential technical fact for the epigraphic record examined in Parts Two and Three: because a karaṇa is one event rather than three, a sculptor cannot divide the task across separate panels and must instead solve, within a single static relief, the representational problem of capturing three simultaneous but physically distinct actions (a leg's trajectory, a torso's base, two arms' independent positions) as a single legible unit.
The sthāna component draws on a defined inventory of standing bases, classifiable along two independent axes: weight distribution (samapāda, evenly weighted on both legs, against ayata and aiḍa sthānas, weighted asymmetrically onto one leg) and vertical compression (an upright, minimally bent-knee stance against the deeply bent-knee vaiṣṇava and maṇḍala sthānas associated with a wide, grounded base). These two axes combine to generate the specific sthāna vocabulary each of the 108 karaṇas draws a single member from, and the combination is not arbitrary: a deeply compressed maṇḍala-type sthāna is documented in the commentarial tradition as the physically necessary base for any karaṇa also requiring a vigorous or high-velocity cārī, since the lowered centre of gravity supplies the stability such a cārī requires — meaning the sthāna and cārī components of any given karaṇa are not independently selected but are constrained by the physical requirements the paired cārī imposes, a constraint that becomes directly relevant to Part Four's quartile-by-quartile analysis of the catalogue's own increasing physical intensity.
The Nāṭyaśāstra recognises sixteen cārīs, each a compound of a specific foot position and a specific leg-raising, extending, or sweeping action, divided at the broadest level into two classes: bhaumī cārīs, in which the moving leg remains at or close to floor level throughout the action, and ākāśikī cārīs, in which the leg is raised well above hip height, into the "aerial" register the term ākāśa (sky, space) names directly. This bhaumī/ākāśikī division is the single most consequential classificatory fact governing how a karaṇa is rendered sculpturally: an ākāśikī-cārī karaṇa requires a sculptor to depict a raised leg extending into the upper portion of the panel's own frame, immediately and visually distinguishing such karaṇas from bhaumī-cārī karaṇas even before any finer sthāna or hasta detail is examined — which is precisely why Part Four's quartile analysis is able to track the catalogue's own escalating physical intensity directly through the proportion of ākāśikī to bhaumī cārīs dominant in each successive quartile.
A recurring source of confusion in less careful treatments of the karaṇa system is the conflation of the nṛtta-hasta — the purely formal, non-representational hand and arm position paired with a given sthāna and cārī within a karaṇa — with the very differently functioning hasta-mudrās of abhinaya, the narrative and emotionally expressive hand gestures (śākhā and nṛtya hastas) treated in the Nāṭyaśāstra's separate ninth chapter and used throughout dramatic and expressional performance to denote specific objects, actions, emotional states, or characters. The distinction is not merely terminological. An abhinaya hasta carries independent semantic content legible to a trained audience regardless of what the rest of the body is doing; a nṛtta-hasta carries no such content and functions purely as a formal, ornamental counterpart to the leg and torso action, meaning a karaṇa can be, and typically is, executed without any accompanying narrative facial or gestural abhinaya at all. This is the technical basis for §4's classification of the entire karaṇa system under nṛtta rather than nṛtya or nāṭya, and it is why a temple panel depicting a karaṇa is correctly read as a technical, formal document rather than as a narrative scene requiring a story to be reconstructed from it.
Beneath the sthāna–cārī–hasta formula lies a further, finer layer of distinguishing detail: the recakas, four distinct rotational or circling qualities assignable to the hand, foot, waist, or neck, examined fully in §5 below. A recaka is not a fourth independent component alongside sthāna, cārī, and hasta; it is a modifying quality that the cārī or hasta component of a given karaṇa may or may not incorporate, and it is frequently the presence, absence, or specific type of recaka that differentiates two karaṇas sharing an otherwise identical sthāna and cārī — the combinatorial mechanism, introduced here and developed fully in §5, that allows a comparatively compact underlying vocabulary of components to generate the full, non-redundant 108-item catalogue without requiring 108 entirely independent movement types invented from nothing.
The claim that a karaṇa is indivisible — one event, not three in sequence — is not merely a definitional stipulation; it is testable against the tradition's own compositional practice at the aṅgahāra level, examined in §4. If karaṇas were genuinely divisible into their three components, an aṅgahāra's fixed sequence could in principle be described equally well as a sequence of sthānas, a separate sequence of cārīs, and a separate sequence of hastas, freely recombinable across karaṇa boundaries. The Nāṭyaśāstra's own aṅgahāra definitions do not permit this: they name and sequence karaṇas as whole units, never their sub-components independently, which is the clearest internal textual evidence that the tradition itself treated the karaṇa's tripartite unity as a genuine compositional fact rather than as an analytical convenience introduced later by commentators or modern scholars.
Mārga (path) and deśa (country, region) supply the roots of mārgi and deśī respectively, and the distinction predates and extends beyond dance specifically, appearing across Sanskrit poetics, music theory, and dramaturgy as a general framework distinguishing textually sanctioned, trans-regional classical standard from locally elaborated, vernacular practice. Within the Nāṭyaśāstra's own usage, mārgi practice is associated with the text's own prescriptive authority — the standard the treatise itself lays down as binding wherever its authority is recognised — while deśī practice is associated with the regionally specific elaborations that inevitably develop around any performance tradition as it is transmitted across a subcontinent's worth of distinct courts, temples, and communities, each with their own local aesthetic preferences and occasions of performance.
Of every technical layer within the Nāṭyaśāstra's own vocabulary, the 108 karaṇas are treated with the least regional latitude. Where the hasta-mudrā system (§2.4) is documented as having accumulated substantial regional variation without controversy, and where cārī and sthāna terminology shows some regional divergence in how later texts such as the Abhinaya Darpaṇa restate it (a point developed fully in Part Two), the core 108-item catalogue itself — its number, its named sequence, its status as Śiva's own transmitted dance (§1) — is treated across every commentarial and sculptural source this register draws on as a single, non-regional, mārgi standard. This is the direct consequence of the theological closure argument developed in §1.3: a system presented as a specific, complete, divinely transmitted performance is structurally resistant to the kind of regional elaboration a more clearly human-compiled technical inventory can absorb without doctrinal difficulty.
This does not mean deśī practice played no role in the karaṇa system's own historical life. Regional court and temple performance traditions across South India developed extensive deśī elaboration in the choreographic use of karaṇas — which aṅgahāras were favoured in a given region's own repertoire, how karaṇa sequences were combined with regionally specific abhinaya, musical setting, and occasion — without altering the 108-item catalogue's own core definitions. The deśī layer, in other words, operates at the level of selection, combination, and performance context, while the mārgi layer operates at the level of the individual karaṇa's own definition — a division that allows the tradition to accommodate substantial regional performance diversity while still treating the underlying 108-unit vocabulary as a single, stable, pan-Indian reference standard, exactly the standard the sculptural programs of Part Three each independently claim to be reproducing.
Each of the five temples examined in Part Three is, considered as an institution, a specifically regional foundation: a royal or local devotional undertaking, funded by a specific dynasty (Chola, Vijayanagara-era, or later successor patronage), embedded in a specific regional liturgical calendar, and staffed by musicians and dancers drawn from a specific regional teaching lineage. That a deśī institution of exactly this kind chose, at each of these five sites, to reproduce the mārgi 108-karaṇa standard in permanent stone is itself a historically significant act of self-positioning: it is a deliberate claim to pan-Indian textual authority made by a regional monument, asserting that this particular Tamil temple's own liturgical dance practice is continuous with, rather than a merely local variant of, Bharata's own universally authoritative standard. Reading the sculptural programs of Part Three with this framework in view changes what the panels are understood to be doing: not local artisans freely interpreting a loosely remembered dance tradition, but a specific, historically situated act of textual and theological legitimation, executed in the medium — permanent architectural stone — best suited to making a claim of enduring, non-regional authority visible to every subsequent generation of worshippers.
The mārgi/deśī distinction established here is not merely of antiquarian interest; it recurs, in modified form, in the twentieth-century codification debates surrounding Bharatanatyam's own revival, examined more fully in the continuity discussion of Part Four's closing section. Questions of which regional or lineage-specific variations in adavu execution, hasta usage, or repertoire selection should be treated as authoritative "classical" standard, and which should be treated as legitimate regional or gharana-specific elaboration, replay — in a modern institutional and pedagogical context rather than a temple-epigraphic one — precisely the mārgi/deśī tension this section has traced back to Bharata's own usage, underscoring that the distinction is a durable structural feature of how Indian classical performance traditions negotiate the relationship between textual authority and living regional practice, rather than a concern specific to the ancient period alone.
Classical dramaturgical theory divides performed movement into three categories distinguished by their relationship to meaning: nāṭya, dramatic performance embedded within plot and character, carrying full narrative content; nṛtya, expressional dance conveying rasa (aesthetic sentiment) and bhāva (emotional state) through abhinaya without necessarily being embedded in a continuous narrative; and nṛtta, pure dance, formally organised and rhythmically executed, carrying no narrative or emotional content of its own. The 108 karaṇas belong unambiguously to the nṛtta category — a direct consequence of the nṛtta-hasta distinction established in §2.4 — and this categorical placement is the single most important fact governing how the entire system, textual and sculptural alike, should be interpreted: a karaṇa's aesthetic function is formal and rhythmic, closer in kind to the non-referential aesthetic pleasure produced by pure musical pattern than to the referential, meaning-carrying pleasure produced by narrative drama or expressional abhinaya.
Above the individual karaṇa, the Nāṭyaśāstra enumerates thirty-two aṅgahāras — fixed, named sequences of karaṇas, each aṅgahāra specifying not merely which karaṇas it comprises but the exact order of their execution. The aṅgahāra is the level at which the karaṇa system's combinatorial claim becomes concrete: a closed inventory of 108 minimal units, combined according to fixed and named rules of sequence, generates a further closed inventory of thirty-two larger compositional units, each independently named and independently defined, rather than left to a choreographer's own discretionary combination. This is a two-tier compositional structure directly analogous, at the structural level examined in this series' companion treatment of sphoṭa theory, to the phoneme–word relationship in spoken language — individual minimal units (karaṇas, phonemes) combining under fixed rule into larger, independently meaningful compositional units (aṅgahāras, words) — while remaining, as §4.3 stresses, considerably more closed and non-productive than that linguistic analogy might suggest.
A performer who executes every karaṇa correctly belonging to a given aṅgahāra, but in an order other than the one the text specifies, has not performed a looser or more personally interpreted version of that aṅgahāra; by the tradition's own reckoning, restated consistently across the commentarial sources examined in Part Two, they have performed something else — a different composition, or no legitimate aṅgahāra at all. This sequence-fidelity requirement is what makes the aṅgahāra level directly relevant to the epigraphic record examined in Part Three: where a temple's own panel sequence along a gopuram or vimāna tracks the Nāṭyaśāstra's own textual sequence of karaṇas, the architectural program is not merely displaying 108 isolated postures but is implicitly encoding aṅgahāra-level sequence information into the building's own physical structure, a claim developed with site-specific evidence throughout Part Three.
A further, still higher compositional tier is documented specifically for group choreography: piṇḍībandhas, formations in which multiple performers combine their own individually executed aṅgahāra-level sequences into a single, spatially coordinated group composition — named formations including lotus- and chain-type spatial arrangements requiring every participating performer's own movement to remain coordinated relative to every other performer's, not merely correct in individual isolation. This three-tier hierarchy — karaṇa, aṅgahāra, piṇḍībandha — completes the compositional architecture that governs every subsequent claim this register makes about how a single sculptural panel (typically depicting one karaṇa) relates to the larger compositional and architectural program (typically encoding aṅgahāra-level, and in principle piṇḍībandha-adjacent, sequence) within which it is physically embedded.
A maṇḍala, in this specific technical sense, denotes the larger spatial pattern — diagonal, circular, or cruciform traversal of the performance area — that certain aṅgahāras trace across the stage as they are executed, a category distinct from, though related in name to, the maṇḍala-type sthāna discussed in §2.2. Because this spatial-traversal dimension of the system leaves no direct trace in a single static relief panel, its evidentiary presence in the epigraphic record is necessarily indirect: it is visible, where visible at all, only through the sequential arrangement of karaṇa panels across a gopuram's own spatial and architectural program — a devotee's own physical movement past a sequence of panels standing in, at one remove, for the spatial traversal the original dance itself would have executed, a correspondence examined directly at each site in Part Three.
| Recaka | Body Region | Character of Motion |
|---|---|---|
| Hasta-recaka | Hand / wrist | A circling or rotational action of the hand distinct from the static nṛtta-hasta positions themselves — the dynamic quality a hasta may carry en route to, or away from, a fixed position. |
| Pāda-recaka | Foot | A rotational or circling action of the foot, distinguished from the cārī's own leg-raising or sweeping action by its localisation at the ankle and foot rather than the whole leg. |
| Kaṭi-recaka | Waist | A rotational action of the hip and waist region, among the most visually consequential recakas for sculptural rendering, since it produces the torsional distortion of the torso documented across the "frozen motion" analysis of Part Three. |
| Grīvā-recaka | Neck | A rotational or circling action of the neck and head, the most restrained in physical amplitude of the four but documented in commentary as carrying independent expressive weight even within nṛtta's own non-narrative register. |
The recaka is smaller in scope than a full karaṇa: it names a rotational quality of a single body region rather than a coordinated whole-body configuration, and it functions combinatorially much as a distinctive phonetic feature functions within a phoneme — the smaller articulatory parameter that combines with others to produce a larger, independently meaningful unit. Just as two phonemes may differ by a single distinctive feature while remaining otherwise identical, two karaṇas may differ solely by the substitution of one recaka for another, or by the presence or absence of a recaka, while sharing an otherwise identical sthāna and cārī — the combinatorial economy that allows a comparatively compact set of underlying components (sthānas, cārīs, hastas, recakas) to generate the full, non-redundant 108-item catalogue.
Bhramarī — full-body spinning or turning about the vertical axis — is documented as a distinct and more physically demanding category than any individual recaka, associated concentrated most heavily with the catalogue's fourth textual quartile (treated fully in Part Four, §24). Where a recaka rotates a single body region while the rest of the body remains comparatively stable, bhramarī rotates the entire body as a unit, and is accordingly associated with the deepest, most stable maṇḍala-type sthānas (§2.2) as the necessary base from which such a rotation can be safely executed and controlled. The catalogue's own placement of bhramarī-dominant karaṇas predominantly in its later quartiles is read by this register, consistent with the pedagogical-sequencing argument of §6.3 below, as reflecting a genuine escalation of physical difficulty across the text's own sequence rather than an arbitrary ordering.
Both recaka and bhramarī movements pose an intensified version of the general frozen-motion problem introduced in §2.1: a rotational or circling action is, by definition, a movement extended through time in a way no single static instant can fully capture, and this difficulty compounds as one moves from a localised recaka (affecting a single limb or region) to a full-body bhramarī (affecting the entire figure at once). This theoretical difficulty is the direct antecedent of the specific sculptural conventions documented and analysed in Part Three's dedicated treatment of frozen motion — asymmetric drapery treatment implying momentum, exaggerated torsional distortion beyond any single static pose's natural limit, paired sequential panels borrowing what amounts to proto-cinematic sequential-frame logic — each of which exists specifically to compensate for the representational gap this section identifies at the theoretical level.
Where §4.5 introduced maṇḍala as the spatial trace of an aṅgahāra's own traversal of the performance space, this section situates that category within the full three-level scale of movement the karaṇa system organises: the recaka, operating at the scale of a single limb or body region; the karaṇa itself and its constituent bhramarī-type rotations, operating at the scale of the whole body in place; and maṇḍala movement, operating at the scale of the whole body's traversal of the performance area over the course of a complete aṅgahāra or piṇḍībandha. This three-scale organisation — limb, body, space — supplies the theoretical vocabulary Part Four's catalogue-level analysis depends on throughout, since each of the 108 individual entries can be, and in Part Four is, characterised precisely along all three of these scales simultaneously.
The Nāṭyaśāstra's overall architecture proceeds from an account of dramatic origins and theatre construction, through the theory of rasa and bhāva (Chapters Six and Seven), the hasta-mudrā system (Chapter Nine), and the cārīs and maṇḍalas treated more fully elsewhere (Chapters Ten through Twelve), among a wide range of further technical and theoretical material spanning music, character types, and production practice. Chapter Four occupies an early and structurally distinct position within this architecture: rather than building general dramaturgical theory, it functions as a closed, self-contained technical appendix — 108 named definitions, given in compact verse form, with minimal elaboration and no accompanying dramaturgical commentary within the root chapter itself. This structural distinctness is itself evidence for how the tradition understood the karaṇa catalogue's own status: not as one topic among many within a continuous theoretical exposition, but as a discrete, quasi-independent unit of received material, consistent with the transmission narrative of §1.2, inserted into the larger text as a defined and closed whole.
Each karaṇa definition is compressed to the minimum verse length capable of naming its sthāna, cārī, and hasta components, a style closer to sūtra literature's own extreme compression than to the more expansive descriptive verse found elsewhere in the Nāṭyaśāstra. This terseness is best read, consistent with the broader oral-textual transmission culture the Nāṭyaśāstra belongs to, as evidence that the root verses were designed to function as memory-retrieval cues for a reader or reciter who already possessed the underlying physical vocabulary through direct bodily training, rather than as self-sufficient written instructions adequate for a reader with no prior physical grounding — a point developed with its full evidentiary consequences in Part Two's treatment of the commentarial tradition Abhinavagupta's Abhinavabhāratī supplies to close exactly this gap.
Chapter Four's placement early in the Nāṭyaśāstra's overall sequence, preceding the more expansive treatment of abhinaya and expressional performance in later chapters, carries an implicit theoretical claim about pedagogical and logical priority: pure kinetic form (nṛtta) is treated as prior to expressional dance (nṛtya) and dramatic performance (nāṭya), both in the text's own expository order and, by direct inheritance, in the training progression standard across later classical dance pedagogy, in which a student's early training is given substantially to pure-movement and adavu-level work before expressional repertoire requiring narrative and emotional content is introduced. This sequencing is not merely a pedagogical convenience; it follows directly from the categorical analysis of §4.1, since a student cannot meaningfully be taught to layer narrative and emotional content (nṛtya, nāṭya) onto a physical vocabulary (nṛtta) they have not yet acquired.
Because the karaṇa system belongs categorically to nṛtta rather than nṛtya or nāṭya (§4.1), this register maintains a consistent methodological boundary throughout Parts Two through Four: it does not attempt to read narrative or emotional content into individual karaṇas or their sculptural depictions, even where a given panel's pose might, considered in isolation and without reference to its textual definition, invite such a reading. A karaṇa's correct interpretive register is formal and technical, and the later Nāṭyaśāstra chapters treating hasta-mudrā, bhāva, and rasa — while indispensable to understanding Indian classical performance as a whole — belong to a categorically distinct portion of the text, engaged by this register only where directly necessary (as in §2.4's disambiguation of nṛtta-hasta from abhinaya hasta) and not treated as continuous source material for interpreting the 108 karaṇas themselves.
The Catalogue of 108 by Textual Quartile, and Its Living Continuity in Classical Dance.
अष्टोत्तरशतकरणानि — शिला च शरीरम्Bharata's fourth chapter does not present the 108 karaṇas as an undifferentiated list; the catalogue has an internal shape, moving from simpler, primarily standing-based configurations toward increasingly vigorous rotational, aerial, and combination movements as it proceeds. This volume follows that shape directly, organising the full catalogue into four textual quartiles of twenty-seven karaṇas each — the division this register uses throughout, bounded at Talapuṣpapuṭa, Vikṣiptākṣiptaka, Ardharecita, Bhujaṅgāñcitaka, Daṇḍarecita, Valitaka, Ghūrṇita, and Vaiśākha — and reads each quartile against its own dominant sthāna-cārī-hasta logic, its own characteristic recaka and bhramarī content, and its own epigraphic distribution across the five sites established in Volume I.
Before the catalogue itself can be read, its three constituent sub-systems require their own definition, since every entry in §21 through §24 is stated as a specific combination of one member from each. The sthāna system comprises the standing bases from which a karaṇa departs and to which, in many cases, it returns — including the samapāda (even, weight-balanced) sthānas, the vaiṣṇava and maṇḍala sthānas involving deep knee-bend and wide stance, and the aiḍa and ayata sthānas involving asymmetric weight distribution onto one leg. The cārī system, sixteen in number in the root text, combines a specific foot-position with a specific leg-raising, extending, or sweeping action, divided broadly into bhaumī cārīs (grounded, floor-level leg actions) and ākāśikī cārīs (aerial or elevated leg actions involving the leg raised well above hip height) — a division that correlates directly with the visual vigour of the karaṇas built from them, and one immediately legible in the sculptural record, where an ākāśikī-cārī karaṇa is typically rendered with a leg raised above waist height against the panel's own vertical frame, while a bhaumī-cārī karaṇa keeps both legs within a lower register of the same frame.
The nṛtta-hasta system, finally, comprises the purely formal, non-narrative hand and arm positions assigned to accompany each sthāna–cārī pairing — recakita, ardharecita, and related terms describing the degree and direction of the arm's own circling or extension — and it is this component, more than the sthāna or cārī, that supplies the fine visual differentiation a sculptor relies on to distinguish two karaṇas that otherwise share the same base and leg action, since the arm positions vary more combinatorially than the comparatively smaller sthāna and cārī inventories permit on their own.
The catalogue's opening quartile is dominated by karaṇas built on samapāda and closely related even-weighted sthānas, combined predominantly with bhaumī (grounded) cārīs and comparatively restrained nṛtta-hasta positions. This is not incidental ordering: the sequence functions pedagogically as well as textually, introducing the performer to the fundamental standing bases before the catalogue proceeds, in its later quartiles, to the more vigorous aerial and rotational vocabulary built upon this foundation. The opening karaṇa, Talapuṣpapuṭa — combining a cupped, blossom-like hand configuration with a comparatively simple standing base — is documented across all five sites examined in Volume I as one of the most consistently identified panels in the entire corpus, functioning as an anchor entry against which the triangulated identification method is calibrated before being extended to more ambiguous entries later in the sequence.
| No. | Karaṇa | Component Logic | Epigraphic Correlate |
|---|---|---|---|
| 1 | Talapuṣpapuṭa | Cupped, blossom-form nṛtta-hasta over an even, samapāda base; the catalogue's own opening invocatory posture, read by commentators as an offering-gesture in formal rather than narrative register. | ChidambaramTanjavur Securely labelled at Chidambaram's eastern gopuram; treated as the anchor identification for the full sequence. |
| 5 | Nūpuragraha? — corrected: samapāda-anchored group | An even-base configuration in which the nṛtta-hasta is drawn close to the ankle ornament, associated in commentary with a low, contained arm circling rather than extension. | Chidambaram Identified with moderate confidence; base and hasta components secure, cārī reading dependent on Abhinavagupta's gloss. |
| 10 | Vṛścikakuṭṭita | An early appearance of the vṛścika (scorpion) motif that recurs and intensifies through later quartiles: a curled, raised-leg action against a stable grounded base, prefiguring the more extreme vṛścika-type karaṇas of the second and third quartiles. | Tanjavur Identified on the vimāna's lower tier; one of the earliest-appearing scorpion-motif panels in the Chola sequence. |
| 16 | Bhramaraka (early form) | A restrained, partial rotational action of the kaṭi-recaka type, introducing the recaka vocabulary of §5 within an otherwise grounded, low-intensity sthāna context. | Chidambaram Panel condition partial; identification relies substantially on textual-commentarial triangulation rather than sculptural clarity. |
| 27 | Vikṣiptākṣiptaka | Closing entry of the first quartile: a thrown-and-drawn-back leg action (kṣipta, "thrown/cast," doubled in the compound) combined with a correspondingly extended hasta, marking a deliberate transition in intensity toward the second quartile's more actively rotational vocabulary. | ChidambaramTiruvannamalai Positioned at the close of the first gopuram tier at both sites, consistent with its textual role as a quartile boundary. |
The second quartile is where the recaka vocabulary of §5 becomes the catalogue's dominant organising principle: a substantial proportion of these twenty-seven entries are built around a named, partial (ardha-, "half") rotational action — ardharecita itself opens the quartile — combined with sthānas that begin to introduce asymmetric, single-leg weight distribution rather than the even samapāda bases dominant in the first quartile. This is also the quartile in which the bhujaṅga (serpent) motif enters the catalogue's own imagery in its named terminology, closing the quartile with Bhujaṅgāñcitaka — a curved, serpentine torso-and-arm configuration whose sculptural rendering, across every site in this register, is among the most visually distinctive and least ambiguous in the entire 108, owing to the pronounced lateral torso curve the pose requires, a curve no other karaṇa in the catalogue produces in quite the same register.
| No. | Karaṇa | Component Logic | Epigraphic Correlate |
|---|---|---|---|
| 28 | Ardharecita | A half-turn of the kaṭi-recaka combined with an ayata (single-leg-weighted) sthāna; the quartile's own opening entry, establishing the partial-rotation logic that recurs through the following twenty-six karaṇas. | Chidambaram Clearly labelled; torso torsion legible even on weathered sections of the eastern gopuram. |
| 33 | Latāvṛścika | A creeper-like (latā) extension of the leg combined with the scorpion (vṛścika) curling motif introduced in the first quartile's Vṛścikakuṭṭita, now intensified into a fuller aerial extension. | Tanjavur Identified on the vimāna's mid-tier; among the panels Subrahmanyam's research first connected across the Chidambaram–Tanjavur comparative sequence. |
| 40 | Daṇḍapakṣa | A rod-straight (daṇḍa) extended leg combined with a wing-like (pakṣa) spread of the nṛtta-hastas; associated in commentary with an ākāśikī cārī requiring the raised leg held at or above hip height. | ChidambaramKumbakonam Present in both the gopuram sequence and the Sarangapani facade register, with minor stylistic divergence in hasta angle. |
| 46 | Kaṭicchinna | A pronounced lateral break (chinna, "cut/severed") at the waist, combined with a stabilising grounded base; one of several karaṇas in this quartile whose name directly names the visual effect a sculptor must render, easing identification. | Tanjavur Securely identified; the waist-break is among the more legible torsional features even at the vimāna's considerable carved height. |
| 54 | Bhujaṅgāñcitaka | Closing entry: a full serpentine (bhujaṅga) curve of the torso combined with an āñcita (bent, drawn-in) leg action, marking the quartile's most visually intense configuration and the transition point into the third quartile's own combination-karaṇa vocabulary. | ChidambaramTanjavurTiruvannamalai The single most consistently and confidently identified karaṇa across all three major gopuram/vimāna sites after Talapuṣpapuṭa itself. |
The third quartile is characterised by an increase in compound and combination karaṇas — configurations in which two distinct component-actions established separately in the first two quartiles (a daṇḍa-type extended limb, a recaka-type rotation, a vṛścika-type curl) are combined within a single karaṇa rather than appearing individually. This compounding is textually signalled by the quartile's own vocabulary of compound technical terms, and it corresponds, in the sculptural record, to a marked increase in the physical complexity of the poses the workshops at each site were required to render — a complexity reflected in a measurably higher rate of ambiguous or contested identification among researchers for this quartile than for the first two, since a compound karaṇa's sculptural rendering can, in a worn or damaged panel, be mistaken for a related but textually distinct compound entry elsewhere in the same stretch of the catalogue. The quartile closes with Valitaka, a turned, twisting configuration whose name (from vali-, "to turn/twist") signals the transition toward the fourth quartile's dominant rotational and bhramarī vocabulary.
| No. | Karaṇa | Component Logic | Epigraphic Correlate |
|---|---|---|---|
| 55 | Daṇḍarecita | A rod-straight extended limb combined with a full, rather than partial, recaka rotation — distinguishing it from the second quartile's Ardharecita by the completeness of the rotational component. | Chidambaram Identification depends on distinguishing full versus partial recaka in a weathered panel; Abhinavagupta's gloss is the deciding evidence in the standard reading. |
| 61 | Nikuṭṭaka | A stamping, percussive foot-action (kuṭṭ-, "to strike/pound") combined with a contained, close-held hasta pair; associated in the tradition with an audible rhythmic accent marking a specific tāla position within its aṅgahāra. | TanjavurKumbakonam Present at both sites; the percussive foot-strike is rendered through a distinctive raised-heel convention shared across both workshops. |
| 68 | Ardhanikuṭṭaka | The half-form counterpart to Nikuṭṭaka above, reducing the percussive strike to a partial, restrained action — an instance of the paired full/half naming convention (recita/ardharecita; nikuṭṭaka/ardhanikuṭṭaka) that recurs structurally across the catalogue. | Chidambaram Identified by direct comparison against the adjacent, better-preserved Nikuṭṭaka panel. |
| 75 | Vṛṣabhakrīḍita | A "bull's-play" (vṛṣabha-krīḍita) configuration involving a wide, grounded stance with a pawing or striking leg-action, associated in commentary with a specific, more percussive cārī than the aerial ākāśikī cārīs dominant elsewhere in this quartile. | Tanjavur One of the more physically distinctive compound karaṇas in the vimāna sequence, aiding secure identification despite the panel's considerable height. |
| 81 | Valitaka | Closing entry: a twisting (vali-) rotation of the full torso against a stabilising base, functioning textually and iconographically as the pivot into the fourth quartile's intensified bhramarī vocabulary. | ChidambaramTiruvannamalai Positioned at a tier boundary at both sites, consistent with its role as a structural quartile marker. |
The catalogue's final quartile is dominated by the bhramarī (spinning/turning) vocabulary introduced structurally in §5, and by the most physically demanding ākāśikī cārīs in the entire 108 — full-body rotations, elevated leg extensions well above hip height, and, in several documented entries, configurations requiring a controlled loss and recovery of vertical alignment, of which Ghūrṇita ("whirling, giddy") is the opening and most direct textual example. This is also the quartile in which the "frozen motion" problem discussed in Volume I, §18 becomes most acute, and where the secondary sculptural convention described there — asymmetric drapery treatment, exaggerated torsional distortion, paired sequential panels — appears with its highest concentration across the five sites, precisely because a single static instant is least able, on its own, to convey a full-body spin without some such compensating device. The catalogue closes with Vaiśākha, a named, wide, symmetrically balanced stance whose comparative visual stability — after twenty-six preceding karaṇas of escalating rotational and aerial intensity — functions as a deliberate textual and choreographic resolution, returning the performer to a grounded, controlled base at the sequence's own close, structurally answering the opening quartile's own grounded Talapuṣpapuṭa.
| No. | Karaṇa | Component Logic | Epigraphic Correlate |
|---|---|---|---|
| 82 | Ghūrṇita | A full rotational turn of the body about its own vertical axis, opening the quartile's bhramarī vocabulary directly; commentary specifies the turn as continuous rather than the partial (ardha-) rotations of earlier quartiles. | Chidambaram Rendered via the exaggerated-torsion convention of Vol. I §18; drapery treatment implies rotational momentum. |
| 89 | Alātaka | Named for the alāta, the firebrand or torch whose circular swing traces a visible ring when moved quickly — a karaṇa built around a wide, sweeping ākāśikī leg action evoking that same circular after-image. | TanjavurTiruvannamalai The wide leg-sweep is among the most legible ākāśikī renderings in the corpus, aiding confident cross-site identification. |
| 95 | Nūpuragrahaṇa? (elevated form) | An elevated variant building on the grounded ankle-focused vocabulary of the first quartile, now executed from an ākāśikī rather than bhaumī cārī base — the catalogue's own compositional habit of revisiting earlier motifs at higher intensity, discussed structurally at the head of this quartile. | Kumbakonam Identification supported by comparison against the corresponding lower-intensity first-quartile motif on the same facade register. |
| 101 | [Boundary of the Vṛddhācalam sequence] | The Vṛddhagiriśvara program (Vol. I, §17) terminates at this point in its own surviving panel sequence; karaṇas 102–108 are absent from that site specifically, a gap addressed directly in Volume I's discussion of the temple's incomplete commission. | Vṛddhācalam Final securely surviving panel in the site's own sequence. |
| 108 | Vaiśākha | Closing entry of the entire 108-karaṇa catalogue: a wide, symmetrical, evenly weighted stance, deliberately contrasting the preceding bhramarī-dominated stretch of the fourth quartile and functioning as the sequence's own structural resolution. | ChidambaramTanjavur Positioned at the terminal point of both major sequences; its comparative stability makes it, alongside Talapuṣpapuṭa, one of the two most confidently identified panels in the entire corpus. |
The karaṇa system's afterlife in living practice is neither uniform nor complete across the classical forms that claim descent from Bharata's text, and the honest art-historical account distinguishes carefully between direct structural continuity, partial vocabulary survival, and modern reconstructive revival, rather than presenting all three as equivalent. Bharatanatyam's own foundational adavu vocabulary — the basic combinatorial units of araimaṇḍi (half-seated base), foot strikes, and arm positions that every student learns before advancing to full compositional repertoire — is widely read by dance scholars as a structurally descended, though not verse-for-verse identical, elaboration of the sthāna–cārī–hasta logic §20 describes, transmitted through the guru-śiṣya paramparā of the naṭṭuvanār teaching lineages associated historically with several of the temples examined in Volume I, rather than through direct, continuous, verse-by-verse study of the Nāṭyaśāstra's own fourth chapter, which had itself fallen into substantial practical obscurity by the colonial period before the twentieth-century revival movements associated with figures including Rukmini Devi Arundale restored it to active scholarly and pedagogical attention.
Kuchipudi's own historical development, rooted in a distinct Andhra devotional and dance-drama tradition rather than in the Tamil temple-sculptural corpus this register centres on, preserves a related but independently evolved relationship to the same Nāṭyaśāstra source material — sharing Bharata's underlying sthāna and cārī vocabulary at a structural level while having developed its own distinct choreographic conventions, including the tarangam and specific narrative sequences not paralleled in the karaṇa-focused pure-nṛtta tradition this register documents. It is specifically Padma Subrahmanyam's own twentieth-century reconstructive research program (Vol. I, §12) that constitutes the most direct and deliberate act of continuity between the epigraphic record and contemporary performance: her identification of specific temple panels with specific numbered karaṇas was undertaken not as a purely archival exercise but explicitly in order to make those identified karaṇas physically re-performable, and her own subsequent choreographic and pedagogical work — developing what she and her students term Bhāratanṛtyam — represents a documented, traceable case in which stone carved eight to nine centuries earlier was read back into a living dancer's trained body, closing, for these specific identified karaṇas at least, the very gap between text, stone, and flesh this entire register has been organised to examine.
What should not be overstated, in closing, is the completeness of this continuity. The overwhelming majority of contemporary Bharatanatyam and Kuchipudi repertoire is not composed of directly reconstructed, individually identified karaṇas performed in their original Nāṭyaśāstra sequence; it is composed of choreography developed within each tradition's own historically evolved compositional conventions, informed by but not reducible to the karaṇa catalogue examined in this Part. The epigraphic and reconstructive record establishes, with considerable confidence, that the 108 karaṇas were once a living, performed vocabulary, textually defined, commentarially glossed, and monumentally inscribed in stone at five specific Tamil temples across several centuries of continuous royal and devotional patronage — and that a documented, scholarly, and pedagogically active bridge now exists, through Subrahmanyam's and subsequent researchers' work, between that historical record and a present-day dancer's own trained body. It does not establish an unbroken, continuous, century-by-century performance tradition of the complete 108 in their original textual sequence, and this register closes by stating that distinction plainly, in keeping with the evidentiary discipline maintained throughout both volumes.
Chidambaram, Tanjavur, Kumbakonam, Tiruvannamalai, and Vṛddhācalam read as a single comparative sculptural corpus, together with the representational problem stone poses for a system defined as motion.
कलेतिहास-मूर्तिविज्ञान-विवरणम्Five Tamil temples, spanning roughly four centuries of continuous royal and local patronage across Chola, Vijayanagara-successor, and later regional building campaigns, each independently committed the 108-karaṇa catalogue to permanent stone. No two committed it in the same way. This Part reads each site on its own architectural and iconographic terms before drawing the comparative and theoretical conclusions — about frozen motion, about the iconography of the dancing god, about what a sculptural workshop does when asked to carve an event rather than a form — that only become visible once all five are read together as a single corpus rather than as five isolated monuments.
The Chidambaram Naṭarāja temple's karaṇa program is carved principally across the eastern gopuram of the Naṭarāja Sabhā precinct, with a secondary, less completely preserved sequence on the western gopuram. The eastern gopuram's own construction is conventionally dated to the Pandya-period expansion of the temple complex, considerably later than the temple's own Chola-period core, which matters directly for how the sequence should be read chronologically: the karaṇa panels are not contemporaneous with the earliest phase of temple construction but belong to a specific, datable campaign of gopuram-building undertaken with the explicit purpose of monumentalising the dance vocabulary at the temple's own ceremonial threshold, a purpose distinct from and later than the establishment of the sanctum itself.
Each karaṇa at Chidambaram occupies an individually framed niche within the gopuram's own tiered structure, a physical unit consistent enough in proportion across the full sequence that the niche itself functions as a kind of implicit grid, holding one kara�knoa's worth of sthāna, cārī, and hasta information within a fixed and repeated architectural frame. This consistency of framing is part of what makes the sequence legible as a catalogue rather than as an ornamental frieze: a viewer moving along the gopuram's tiers encounters the same visual unit — one niche, one figure, one identified movement — repeated with the same discipline the root text itself imposes at the level of the individual verse.
The presence of an identifying Grantha-script label accompanying a substantial number of the eastern gopuram's panels is, within the five-site corpus this register treats, unique to Chidambaram, and it is this feature specifically that permits the site to function as the calibration base for the triangulated identification method described in Part Two, §12. Where a label survives in legible condition, the identification of the panel requires no recourse to Abhinavagupta's commentary or to comparative sculptural reasoning; the epigraphic record settles the question directly, and it is only where a label has weathered past legibility that the fuller triangulated method must be invoked even at this best-documented of the five sites.
Because Chidambaram is, in Śaiva theology, understood as the specific site of Śiva's own Ānanda Tāṇḍava, the gopuram's karaṇa program cannot be read merely as a didactic display of technical dance vocabulary attached, more or less incidentally, to a major temple. It is read by the temple's own theological tradition, developed at length in the Cidambara Māhātmya and associated Śaiva Siddhānta commentary, as an architectural restatement, at the precinct's own ceremonial threshold, of the specific transmission narrative in which Śiva's own dance becomes the textual and sculptural inheritance a devotee physically passes through en route to the sanctum. The gopuram is, in this reading, not a gateway decorated with dance imagery but a gateway constituted by it.
The western gopuram preserves a less complete and, at several points, differently ordered sequence than the eastern gopuram, a divergence that current scholarship treats as most plausibly attributable to differential preservation and to distinct building campaigns rather than to a deliberate alternative textual ordering. Because the eastern sequence is both more complete and more consistently labelled, this register follows the standard scholarly practice of treating it as the primary reference sequence for the site as a whole, citing the western gopuram's own panels chiefly as corroborating or supplementary evidence where the eastern sequence's own panels are damaged or ambiguous.
Not every niche in the eastern gopuram's own sequence survives in a condition permitting confident identification even with the aid of a legible label; erosion, historical whitewashing campaigns, and structural repairs undertaken across the temple's long continuous history of active worship and maintenance have each, at specific points along the sequence, degraded individual panels beyond the point at which their own sthāna, cārī, and hasta components can be read with the confidence available elsewhere in the corpus. This register follows Subrahmanyam's own documented practice of flagging such panels explicitly rather than supplying a confident identification the physical evidence does not itself support.
The Chidambaram temple complex is organised around two principal ritual halls, the Cit Sabhā (the "hall of consciousness," housing the bronze Naṭarāja and the aniconic ākāśa-liṅga representing Śiva as formless space) and the Kanaka Sabhā immediately adjacent to it. The eastern gopuram's own karaṇa sequence is positioned along the devotee's approach toward this Sabhā complex rather than toward an outlying shrine, which is itself evidence for how centrally the temple's own architects regarded the dance program: it is not peripheral ornament on a subsidiary structure but is aligned with the axis of approach to the temple's own most theologically significant space.
Measured panel-to-panel, the eastern gopuram's niches maintain a consistency of proportion — height-to-width ratio, relief depth, the framing border separating one niche from the next — that is markedly tighter than the corresponding consistency documented at Kumbakonam's facade program in §15, a difference this register attributes to Chidambaram's own gopuram program having been executed, on current art-historical assessment, within a more tightly bounded single building campaign than the Sarangapani facade's own longer and more episodic construction history.
Traces of pigment surviving in protected recesses of several eastern gopuram niches indicate that the karaṇa panels were, at some point in the temple's own history, painted rather than left as bare relief — a common practice across South Indian temple sculpture generally but one whose extent and chronology at Chidambaram specifically remains incompletely documented. Where such pigment traces survive, they are consistent with, though not conclusive proof of, a later repainting campaign distinct from the panels' own original Pandya-period carving, and this register notes the pigment evidence as a live area of ongoing conservation study rather than a settled point.
Because Chidambaram's own inscriptional labels supply the only site-internal confirmation of karaṇa names available anywhere in this five-site corpus, the Sanskrit terminology this register uses throughout Part Four to name individual karaṇas is, wherever a Chidambaram label survives, drawn directly from that label rather than reconstructed solely from the root text or from Abhinavagupta's commentary — a methodological dependency that makes Chidambaram not merely one site among five but the terminological backbone underlying every other site's own comparative treatment in this Part.
The Brihadīśvara temple's construction under Rājarāja I Chola, completed in the early eleventh century, is among the most securely dated of the Chola dynasty's major architectural undertakings, its own inscriptional record supplying a chronology considerably more precise than is available for several of the other sites this register treats. The karaṇa cycle's own placement on the vimāna — the pyramidal superstructure directly over the sanctum, rather than on a subsidiary gopuram — reflects the temple's own overall architectural emphasis: unlike Chidambaram, where the gopurams function as the ceremonial threshold structures a devotee physically passes through, Brihadīśvara's principal architectural statement is the vimāna itself, and the decision to carve the karaṇa sequence there rather than on a gateway structure positions the dance program at the temple's own architectural and theological centre rather than at its threshold.
The vimāna's considerable height — the reliefs are carved at a level well above the vantage point from which the temple's other major sculptural programs can be readily examined — is directly responsible for the karaṇa cycle's long absence from systematic art-historical documentation prior to the twentieth century. Earlier surveys of the temple's sculptural program, concerned chiefly with the more accessible narrative and iconographic reliefs at lower registers, did not undertake the sustained close-range documentation the vimāna's upper tiers required, and it was specifically Padma Subrahmanyam's own research program, applying the triangulated identification method of Part Two, §12 to photographic and direct on-site documentation obtained at the necessary elevation, that first established the cycle's own connection to the Nāṭyaśāstra's textual sequence in systematic form.
The karaṇa panels at Tanjavur are distributed across multiple ascending tiers of the vimāna's own structure, with the panel density per tier governed by the physical proportions of the tier itself rather than by any attempt to preserve a fixed number of karaṇas per architectural register — a constraint the sculptural workshop shares with, though resolves somewhat differently than, the tier-boundary negotiation documented at Tiruvannamalai's own Rāja Gopuram in §16. Lower tiers, closer to the main sanctum's own roofline, tend to preserve the sequence's earlier, more grounded first-quartile karaṇas, while the upper tiers carry a disproportionate share of the more vigorous third- and fourth-quartile bhramarī and ākāśikī vocabulary, a distribution this register reads as broadly, though not with perfect consistency, tracking the textual sequence's own escalation of physical intensity.
Systematic comparison between the Tanjavur and Chidambaram sequences, undertaken across Subrahmanyam's own comparative analysis, documents specific points of divergence in panel ordering that cannot be attributed to differential preservation alone. These divergences are treated by this register as significant evidence for the degree of interpretive latitude a Chola-period sculptural workshop retained even while working, in both cases, from what was understood as the same authoritative mārgi textual source (Part One, §3.2) — evidence that the pan-Indian textual standard the karaṇa catalogue represents did not translate into perfectly uniform architectural execution even among royal workshops operating within the same dynasty and within a comparatively narrow historical window.
Because the Brihadīśvara temple's own donative and dedicatory inscriptional corpus is unusually rich and securely dated relative to Rājarāja I's own regnal chronology, the vimāna's karaṇa cycle benefits from a firmer independent dating anchor than is available for several of the other sites this register treats, allowing the cycle to be positioned with confidence as either directly contemporaneous with, or shortly preceding, the main phase of Chidambaram's own eastern gopuram construction — a relative chronology directly relevant to assessing which site's sequence more plausibly represents the earlier or more textually conservative of the two major Chola-period programs.
The vimāna's elevation, which protected the karaṇa cycle from the kind of ground-level wear, whitewashing, and repeated ritual handling documented at Chidambaram's own more accessible gopuram niches, has not spared the panels from weathering exposure or from the structural stresses inherent to a millennium-old stone superstructure; several panels in the upper tiers preserve their broad sthāna and cārī outlines clearly while losing the finer recaka and hasta detail this register's identifications depend on, a pattern of partial rather than uniform preservation this register notes wherever it bears directly on the confidence of a specific identification.
The Brihadīśvara vimāna carries, beyond the karaṇa cycle itself, an extensive independent program of narrative and iconographic relief spanning multiple registers below the karaṇa tiers, including scenes drawn from Śaiva mythology unconnected to the tāṇḍava narrative specifically. The karaṇa cycle's own placement above rather than among this broader narrative program reinforces the reading, developed in §14.1, that the dance sequence occupies a distinct, formally separated register within the vimāna's overall design — closer in status to an independent technical inset than to a continuation of the surrounding narrative frieze.
The temple's own inscriptional corpus records, independently of the karaṇa cycle itself, Rājarāja I's formal endowment of a substantial body of temple musicians and dancers attached to Brihadīśvara's own ritual calendar, a documented patronage pattern directly consistent with the reading developed in Part Two, §7.3–7.4 for Chidambaram: the sculptural program is best understood as accompanying, rather than substituting for, an actively maintained institutional practice of temple dance contemporaneous with the vimāna's own construction.
The vimāna's own load-bearing structure, tapering progressively toward its summit, imposes physical constraints on niche size and spacing that differ tier by tier in ways a flat gopuram facade does not encounter, and several of the panel-sequence irregularities documented in §14.4 are, on current assessment, at least partly attributable to this structural tapering rather than to compositional choice alone — a further instance of the general principle, developed fully at Tiruvannamalai in §16.4, that architectural constraint and textual fidelity exist in continuous negotiation across every site in this corpus rather than being freely reconcilable at will.
Because Brihadīśvara was commissioned directly by the reigning Chola emperor as an imperial rather than a purely local devotional undertaking, its karaṇa cycle supplies this register's clearest instance of the mārgi-claim argument developed in Part One, §3.4 operating at the scale of dynastic rather than merely local patronage: an emperor's own principal temple reproducing, in its own most prominent architectural register, a textual standard whose authority claim is explicitly pan-Indian and non-regional, a choice of monumental emphasis this register reads as a deliberate imperial statement of textual and cultural legitimacy rather than as an incidental local practice.
The Sarangapani temple's own status as a major Vaiṣṇava institution, dedicated to Viṣṇu rather than Śiva, gives its karaṇa program a distinct art-historical significance independent of its own sculptural quality: the decision to carve the full 108-karaṇa sequence on the facade of a Vaiṣṇava temple demonstrates that by the period of the facade's own construction, the karaṇa catalogue had become sufficiently detached from the specifically Śaiva tāṇḍava mythology examined in Part One, §1 to function as a general technical and ornamental program available to any major devotional institution, Śaiva or Vaiṣṇava alike, without requiring a Naṭarāja-specific theological justification of the kind anchoring the Chidambaram program so directly.
Unlike the gopuram-tier or vimāna-tier placement documented at Chidambaram, Tanjavur, and Tiruvannamalai, the Sarangapani karaṇa panels are carved along the temple's own external facade at a single, or in places double, horizontal register maintained at a broadly consistent viewing height throughout. This placement produces markedly different viewing conditions than the vertically ascending programs at the other sites: a devotee historically encountered the Sarangapani sequence as a continuous horizontal band, legible in a single circumambulatory pass around the temple's own exterior, rather than as a sequence requiring progressively upward attention as at a multi-tier gopuram.
The horizontal, single-register arrangement carries a specific methodological advantage for modern sequence-mapping: because the panels are not distributed across multiple architecturally discontinuous tiers, the question of which panel follows which along the sequence is considerably less ambiguous at Kumbakonam than at a multi-tier gopuram program, where tier boundaries and the transition from one architectural register to the next can introduce genuine uncertainty about the intended reading order. This relative clarity of sequence, however, is offset by generally more advanced weathering of the facade reliefs relative to the more sheltered gopuram niches at Chidambaram, a trade-off this register notes explicitly wherever it affects the confidence of a specific Kumbakonam identification.
Comparative stylistic analysis of the Sarangapani panels against the Chidambaram and Tanjavur sequences reveals a distinct regional sculptural idiom in the treatment of drapery, facial modelling, and the proportion of figure to niche, consistent with the deśī-workshop variation the mārgi/deśī framework of Part One, §3.4 predicts: a shared mārgi textual mandate — the same 108-item catalogue, understood as pan-Indian and non-regional in its core definitions — executed through a recognisably distinct regional workshop tradition, itself a further instance of the same pattern documented for panel-sequence divergence between Chidambaram and Tanjavur in §14.4.
The facade arrangement's own correspondence with the standard Hindu temple practice of pradakṣiṇā, ritual circumambulation of the sanctum, gives the Kumbakonam program a devotional function distinct from the threshold-crossing logic of Chidambaram's own gopuram: rather than being physically passed through once, as at a gateway, the Sarangapani sequence is potentially encountered repeatedly, across each circuit of the temple a devotee's own regular worship practice would involve, a repeated exposure this register treats as relevant context for understanding how such a technical catalogue could remain, over centuries, a familiar visual presence within a community's ordinary devotional routine rather than a monument encountered only once.
Unlike the documented and diagnostically significant gap at Vṛddhācalam examined in §17, the Sarangapani facade sequence is not currently documented as terminating short of the full 108; where individual panels are missing or illegible, the pattern is consistent with localised weathering and structural repair rather than with the kind of systematic truncation the Vṛddhācalam case presents, though this register notes that the facade's own more advanced general weathering makes a fully confident claim of completeness harder to establish here than at the better-preserved Chidambaram sequence.
The karaṇa register sits within a facade otherwise dominated by conventional Vaiṣṇava narrative and iconographic relief — avatāra sequences, scenes from Viṣṇu's own principal mythology — and its own insertion into this surrounding program without any accompanying reframing of the dancer figures as Viṣṇu-associated rather than Śiva-associated supports the reading, developed in §15.1, that the karaṇa catalogue had by this period become sufficiently a shared technical inheritance to be carved without theological adaptation, retained in its original Śaiva-tāṇḍava conceptual frame even within an otherwise thoroughly Vaiṣṇava sculptural program.
The Sarangapani temple's own construction spans multiple phases across several centuries, and the karaṇa facade's own specific dating within that longer history remains less precisely fixed than the securely dated Tanjavur vimāna or the Rājarāja-linked inscriptional corpus discussed in §14.5, a comparative dating uncertainty this register notes as a genuine limitation on how confidently the Kumbakonam program can be placed within the broader chronological sequence of the five sites' own karaṇa-carving campaigns.
Given Kumbakonam's own geographic proximity to Tanjavur, art-historical scholarship has raised, without yet reaching firm consensus, the possibility that the two sites' sculptural workshops shared personnel or at least a common regional training lineage, a hypothesis that would help account for specific stylistic resonances documented between the Sarangapani and Brihadīśvara panels' own treatment of drapery and limb proportion — resonances this register notes as suggestive rather than as an established art-historical fact.
The single-register, ground-adjacent placement that shapes devotional viewing per §15.5 equally shapes the program's modern accessibility for close photographic and epigraphic study: the Kumbakonam sequence is, among the five sites, one of the more straightforward to document systematically precisely because it does not require the elevated vantage points the Tanjavur vimāna and the upper tiers of the Chidambaram and Tiruvannamalai gopurams demand.
The Arunachaleśvara temple's karaṇa program belongs to the Rāja Gopuram, whose principal construction phase is conventionally associated with the Vijayanagara-successor period, placing it chronologically later than both the Chidambaram eastern gopuram and the Tanjavur vimāna cycle, and giving the Tiruvannamalai program a distinct comparative value as evidence for how a considerably later sculptural workshop, operating under a different dynastic patronage and a different regional stylistic idiom, approached the same mārgi textual mandate.
The Tiruvannamalai program is, among the five sites this register treats, the clearest instance of a systematic tier-by-tier architectural arrangement of the full sequence, with the gopuram's own ascending storeys used, in the documented reading Subrahmanyam and subsequent researchers apply, as a physical index for the aṅgahāra-level sequencing discussed in Part One, §4.3: a devotee's own ascent of visual attention up the gopuram correlates, tier by tier, with progression through the textual catalogue's own internal sequence, making the architecture itself into a legible diagram of the aṅgahāra structure rather than merely a support for isolated panels.
The Tiruvannamalai panels display sculptural conventions — a particular treatment of ornament density, a distinct proportion of figure to background relief depth, characteristic Vijayanagara-period facial modelling — that distinguish the program stylistically from the earlier Chola-period sequences at Chidambaram and Tanjavur even where the underlying karaṇa identification is textually identical, a distinction this register treats as further confirmation that the mārgi/deśī framework of Part One, §3 operates as fully across a roughly three-century gap in patronage as it does across the comparatively narrower gap separating the Chidambaram and Tanjavur programs from each other.
Because the Rāja Gopuram's own tier divisions impose a fixed, finite number of available niches per storey, dictated by the tier's own physical dimensions rather than by the textual catalogue's own natural divisions, any point at which the Tiruvannamalai sequence must depart from strict textual order to accommodate the architecture's own boundaries is directly observable in the surviving panel arrangement, supplying concrete evidence for how the sculptural workshop negotiated the tension between architectural constraint and textual fidelity — a negotiation Chidambaram's own gopuram, with its differently proportioned tiers and earlier construction period, resolved through a visibly different set of compromises.
Measured against the Chidambaram reference sequence established in Part Two, §11.3, the Tiruvannamalai program preserves textual order with high fidelity across the majority of its own tiers, departing chiefly at points the architectural tier-boundary evidence of 16.4 can independently account for — a correlation this register treats as strong supporting evidence that the observed departures reflect genuine architectural negotiation rather than an alternative or corrupted textual tradition specific to the Tiruvannamalai workshop.
The Rāja Gopuram remains, among the five sites, one of the more physically accessible for close-range documentation, its lower and middle tiers legible from ground level with the aid of standard photographic equipment, though its uppermost tiers present documentation challenges broadly comparable to those affecting the Tanjavur vimāna's own upper registers, a condition this register notes wherever it bears on the relative confidence of specific upper-tier identifications.
Arunachaleśvara temple is unusual among this register's five sites for possessing multiple major gopurams at different points of its own outer enclosure walls, of which the Rāja Gopuram carrying the karaṇa program is the tallest and most architecturally prominent; the karaṇa sequence's own concentration on this specific gopuram, rather than distribution across the complex's several gateway towers, mirrors Chidambaram's own concentration of its primary sequence on a single gopuram (§13.1) even where, as at Tiruvannamalai, multiple comparable structures were available to a commissioning patron.
While the Rāja Gopuram's broad Vijayanagara-successor dating is well established, attribution to a specific reign within that broader period remains, on current scholarship, less precisely fixed than the Rājarāja I attribution available for Tanjavur, a comparative dating imprecision this register notes rather than resolves, consistent with the evidentiary caution this Part maintains throughout regarding claims the current epigraphic and architectural record does not fully support.
As at Tanjavur, the Rāja Gopuram carries an extensive narrative and iconographic sculptural program beyond the karaṇa sequence itself, distributed across lower tiers and the gopuram's own base registers; the karaṇa panels' consistent restriction to specific, architecturally distinct tiers within this larger program is further evidence, alongside the tier-boundary analysis of §16.4, that the dance sequence was conceived by its own commissioning workshop as a formally bounded technical inset rather than as an undifferentiated continuation of the gopuram's broader decorative scheme.
Because Tiruvannamalai's Rāja Gopuram postdates the Chidambaram and Tanjavur programs by a substantial interval, its own high sequence-fidelity to the textual order, documented in §16.5, supplies this register's strongest available evidence that the mārgi standard governing the karaṇa catalogue's core definitions (Part One, §3.2) retained its binding force across a multi-century, multi-dynasty transmission rather than eroding progressively as the tradition moved further from its own Chola-period sculptural precedents.
The Vṛddhagiriśvara temple's own karaṇa program, though smaller in scholarly profile than Chidambaram's or Tanjavur's, occupies a distinct and methodologically valuable position within this register's five-site corpus precisely because of its incompleteness: a program that stops short of the full 108 supplies a different kind of evidence — evidence about production, patronage continuity, and possible deliberate selection — than a complete sequence can offer on its own.
The surviving sequence at Vṛddhācalam preserves 101 of the full 108 karaṇas, terminating short of Bharata's own closing entries and leaving a documented, currently unresolved gap of seven karaṇas whose absence is not explained by any surviving inscriptional or textual source specific to the site itself, a fact this register states plainly rather than supplying a speculative explanation dressed as established fact.
The first current explanation, and the one most directly consistent with the broader documented pattern of large-scale South Indian temple sculptural programs dependent on continuous royal or local patronage across what could be a multi-generational construction timeline, treats the gap as evidence of an interrupted or abandoned commission — a patron's death, a shift in political fortune, or a reallocation of resources to a different building priority each supplying a plausible, though at Vṛddhācalam specifically unconfirmed, mechanism by which a systematic sculptural program could stop short of its own textual completeness.
The second current explanation treats the gap as reflecting a deliberate, locally specific decision to omit certain karaṇas on ritual or iconographic grounds specific to this temple's own devotional context — a possibility this register does not treat as established, but does not dismiss, since the distribution of the missing seven karaṇas across the catalogue's own sequence, rather than being concentrated uniformly at the sequence's final entries as a simple "construction stopped partway through" explanation would predict, is instead documented as scattered at several distinct points through the full 108-item order.
This scattered rather than terminal distribution is the single most consequential piece of evidence this register can bring to bear on the choice between the two hypotheses of 17.3 and 17.4: an interrupted commission proceeding through the textual sequence in order would be expected to produce a single continuous gap at the sequence's own end, not a gap distributed at several separate points throughout the full catalogue, and the observed pattern accordingly lends somewhat greater support to the possibility of deliberate, selective omission than to a simple halted-construction account, though this register presents that reading as the more probable rather than the conclusively established explanation.
Because the surviving 101 panels can be checked directly against Chidambaram's fuller and inscriptionally labelled reference sequence (Part Two, §11.3), the specific identity of the seven karaṇas absent from the Vṛddhācalam program is recoverable by comparative elimination even without any surviving inscriptional or textual explanation from the site itself — a demonstration of how the five-site corpus functions as a mutually correcting comparative whole, in which a gap at one site becomes legible and diagnostically meaningful only because a fuller sequence survives elsewhere in the same corpus.
Vṛddhagiriśvara temple, though smaller in overall scale and in modern scholarly profile than Chidambaram, Tanjavur, or Tiruvannamalai, belongs to the same broader Chola-successor devotional building tradition, and its karaṇa program's own incompleteness should not be read as reflecting a lesser institutional commitment to the dance catalogue's own theological significance so much as the specific, locally contingent circumstances — patronage interruption or deliberate selection — examined in §17.3 and §17.4.
Where the surviving 101 panels can be stylistically compared against the Chidambaram and Tanjavur sequences, the Vṛddhācalam workshop displays its own regional treatment of drapery and proportion broadly consistent with the mārgi/deśī pattern documented at Kumbakonam in §15.4, further evidence that the deśī-workshop-executing-a-mārgi-mandate pattern established across this register's better-documented sites extends to the comparatively less-studied Vṛddhācalam program as well.
A complete sequence, however well preserved, can only confirm what a fuller sequence elsewhere has already established; an incomplete sequence, precisely because of its incompleteness, generates a distinct class of evidence — about production process, about patronage continuity, about the possibility of deliberate theological selection — unavailable from any of the four complete or near-complete programs this register otherwise treats, which is why Vṛddhācalam occupies a position in this corpus disproportionate to its own smaller scale.
No inscription currently known from the Vṛddhagiriśvara temple directly addresses the karaṇa program's own construction history or explains its termination at 101 panels; this register notes explicitly that a definitive resolution of the interrupted-commission versus deliberate-omission question raised in §17.3–17.5 would require either the discovery of a currently unknown relevant inscription or a more systematic epigraphic re-survey of the site's own donative and dedicatory corpus than has, to date, been undertaken specifically with this question in view.
Every panel examined across §13 through §17 confronts the same underlying difficulty: a karaṇa, by Bharata's own definition (Part One, §2.1), is a coordinated, momentary event in which sthāna, cārī, and nṛtta-hasta occur together within a single count of time, while a stone relief is, necessarily, a single static instant possessing no literal capacity for temporal extension. The five sculptural programs this register treats do not solve this problem uniformly, and the specific solution a given workshop adopts is itself art-historically informative evidence about how that workshop understood the relationship between text and stone.
The convention documented most consistently across all five sites, and most clearly at Chidambaram specifically, selects for depiction the instant of maximum extension or maximum torsional tension within a given karaṇa's own trajectory — the point at which the sthāna, cārī, and nṛtta-hasta components are most clearly and unambiguously differentiated from a neutral standing position. The implicit reasoning, consistent across the sculptural tradition, appears to be that this is the instant from which a viewer already familiar with the movement's own trajectory — whether through direct dance training or through the kind of embodied cultural familiarity a temple's own devotee community could reasonably be expected to possess — can most reliably reconstruct the whole action, both what preceded the depicted instant and what follows it.
A secondary and less common device, documented in a smaller number of panels concentrated particularly among the fourth-quartile bhramarī-type and vṛścika-type karaṇas discussed further in Part Four, achieves a more explicit suggestion of motion through asymmetric drapery treatment: the carved fabric is rendered as trailing or displaced in a direction inconsistent with a figure at rest, implying, through a purely visual convention with no literal counterpart in the dancer's own physical costume at any single instant, the momentum of a body caught mid-rotation rather than posed in stillness.
A related device, documented at several bhramarī-adjacent panels across the corpus, exaggerates the torsional distortion of the torso beyond what any single static pose could naturally hold without loss of balance — a deliberate departure from strict anatomical plausibility undertaken specifically to convey the sensation of a body still in the process of turning, rather than a body that has arrived at and stabilised within a final rotated position, a distinction the sculptor communicates precisely by declining to render the pose as one a real dancer could indefinitely sustain.
The most sophisticated solution documented in this register's corpus involves the deliberate juxtaposition of two adjacent panels depicting sequential moments of what the Nāṭyaśāstra's own aṅgahāra sequencing (Part One, §4.3, §11) indicates were performed as a continuous pair — a device that effectively borrows the sequential-frame logic later associated with cinematic storyboarding, centuries before the concept existed under that name, allowing the temporal extension no single panel can achieve to be reconstructed by a viewer moving physically along the gopuram's own architecture from one niche to the next.
Taken together, the four devices examined in 18.2 through 18.5 support a reading of the sculptural programs, particularly at the sites where sequence-fidelity to the textual order is most fully documented, as functioning collectively closer to an illuminated manuscript's own relationship between text and image than to a gallery of independent decorative postures: each panel is legible in isolation as a single named karaṇa, but the full temporal and choreographic meaning of the tāṇḍava the panels collectively depict is recoverable only by reading the sequence as a sequence, precisely as the aṅgahāra structure of Part One, §4 requires for the dance itself.
Consistent with the categorical boundary established in Part One, §2.4 and §4.1, the facial treatment across nṛtta-register karaṇa panels at all five sites is markedly more restrained than the facial treatment documented in this same temple corpus's own nāṭya- and nṛtya-register narrative reliefs, where dramatic and emotional facial modelling carries genuine expressive weight; this restraint is itself a further, independent line of sculptural evidence — beyond the nṛtta-hasta distinction of Part One, §2.4 — confirming that the workshops themselves understood the karaṇa panels as belonging to a formally distinct representational category from the temples' own narrative sculptural programs.
A small number of panels across the corpus, concentrated among the more grounded first-quartile entries, employ a subtly raised or textured pedestal beneath the standing foot to imply weight-bearing pressure at the instant of the sthāna's own execution — a comparatively modest device set against the more dramatic torsional and drapery-based conventions of §18.3–18.5, but one this register notes as evidence that the frozen-motion problem was addressed through a wider repertoire of sculptural solutions than the two or three dominant conventions alone might suggest.
The maximum-extension convention of §18.2 is documented as the overwhelmingly dominant solution at all five sites and across all four textual quartiles; the more elaborate compensating devices of §18.3 through §18.5 are, by contrast, concentrated disproportionately at Chidambaram and Tanjavur specifically, a distribution this register attributes tentatively to those two sites' own comparatively larger, better-resourced sculptural workshops having greater latitude to attempt more technically demanding representational solutions than the workshops at Kumbakonam, Tiruvannamalai, or Vṛddhācalam.
That five independent workshops, separated by centuries and by regional style, converge overwhelmingly on the same maximum-extension solution to an identical representational problem is itself indirect evidence about how those workshops read the root text's own terse verses: not as instructions for a single static pose to be held, but as compressed descriptions of a trajectory whose most legible single instant a trained sculptor, working in consultation with a practising dance master per Part Two, §9.3, could identify and render with confidence.
Although the 108-karaṇa catalogue is textually and mythologically anchored to Śiva's own tāṇḍava (Part One, §1), the overwhelming majority of panels across all five sites this register treats depict an unidentified or generically rendered male dancer figure, executing the karaṇa in question without the specific iconographic attributes — the flaming halo (prabhāmaṇḍala), the damaru hand-drum, the encircled dwarf-figure Apasmāra beneath the dancer's own foot — that would mark the figure unambiguously as Naṭarāja rather than as a human performer or a gaṇa attendant demonstrating the movement for instructional purposes.
Where a panel is intended to represent Naṭarāja specifically, as documented at certain Chidambaram panels closely associated with the central sanctum icon, the sculptural workshop deploys a recognisable and consistent attribute set — the halo of flame, the damaru, the encircling ring, the trampled Apasmāra figure symbolising ignorance overcome — that is conspicuously withheld from the generic didactic panels making up the bulk of the 108-karaṇa sequence at every site, a withholding this register reads as a deliberate and consistently observed iconographic boundary rather than an inconsistency in workshop practice.
This consistent iconographic restraint carries direct theological weight: it suggests the temples' own sculptural workshops distinguished carefully between the singular, doctrinally load-bearing Naṭarāja icon — reserved for the bronze mūla-mūrti and its principal stone representations, images understood as embodying an actual theophany rather than merely illustrating a technical posture — and the generic didactic figure appropriate to a technical catalogue illustrating the grammar of dance rather than depicting a specific divine manifestation at each of its 108 instances, a distinction directly continuous with the nṛtta/nṛtya/nāṭya categorical boundary examined in Part One, §4.1.
Where a generic dancer figure in the karaṇa sequence is not read simply as an unmarked human performer, art-historical interpretation has proposed reading such figures as gaṇas — Śiva's own attendant beings, frequently depicted in South Indian temple sculpture as participants in and witnesses to the god's own dance — a reading that would preserve a link to the tāṇḍava mythology of Part One, §1 without requiring every individual panel to depict Śiva himself, and would be consistent with the broader iconographic program at several of these temples in which gaṇa figures appear extensively in contexts adjacent to, though distinct from, the central deity's own principal representations.
Where Pārvatī appears within or adjacent to the karaṇa programs examined in this register, her presence is consistently confined to specific narrative or paired panels rather than distributed evenly across the full 108-karaṇa sequence, a pattern directly consistent with the catalogue's own textual attribution to Śiva's tāṇḍava specifically rather than to Pārvatī's own lāsya (Part One, §1.1), and confirming that the sculptural programs, across all five sites, were understood by their own commissioning workshops as illustrating a specific, gendered mythological narrative rather than a generic, ungendered dance-technical inventory available for either deity's own attribution.
Read comparatively, the five sites' own iconographic programs display a consistent core logic — generic dancer figures dominating the technical sequence, specific divine attributes reserved for a small number of theologically load-bearing panels, Pārvatī's presence restricted to distinct narrative contexts — executed through the site-specific regional and period stylistic variations documented individually in §13 through §17, a consistency this register treats as further confirmation that the mārgi textual standard governing the karaṇa catalogue's own core definitions (Part One, §3.2) extended, at these five sites at least, to a broadly shared iconographic discipline as well as to a shared technical vocabulary.
The trampled dwarf-figure Apasmāra, symbolising ignorance overcome and among the most theologically loaded of Naṭarāja's specific attributes, is documented as entirely absent from the generic karaṇa panels at all five sites this register treats, appearing only in the small number of panels already identified in §19.2 as depicting Naṭarāja specifically; this consistent absence across five independent workshops functions as strong corroborating evidence for the iconographic boundary this section has argued for on other grounds, since Apasmāra's presence or absence is among the most visually unambiguous markers separating a theophanic image from a generic didactic one.
Ornament density and costume treatment vary considerably across the generic dancer figures without correlating consistently with any theological distinction this register can identify, suggesting that jewellery and costume elaboration functioned for these workshops chiefly as a marker of the sculptural program's own general decorative register — consistent with a royal or major devotional commission's own visual expectations — rather than as a reliable indicator of which specific figure, divine or human, a given panel was intended to represent.
Where Chidambaram's own Grantha labels name a karaṇa (Part Two, §7.2), they name the movement, not the figure performing it — no surviving label at any of the five sites identifies a generic panel's dancer as Tāṇḍu, as a nameless gaṇa, or as an unspecified human performer, leaving the specific identity question raised in §19.4 formally open at the level of primary textual evidence even where art-historical interpretation has proposed a plausible reading.
The cumulative iconographic evidence examined across this section — restrained attribute use, absence of Apasmāra outside a small number of theophanic panels, restricted and non-uniform placement of Pārvatī, and consistent facial-expressive restraint relative to the temples' own narrative registers — converges on a single, well-supported reading: the 108-karaṇa sculptural programs at all five sites were conceived, executed, and are best interpreted as a formally distinct technical catalogue in stone, theologically anchored to Śiva's own tāṇḍava but iconographically restrained in its individual panels, reserving full theophanic representation for a small number of panels the workshops themselves clearly marked as categorically different from the surrounding didactic sequence.
The full quartile-organised catalogue read against its dominant component logic, its epigraphic distribution across five temples, and its documented afterlife in contemporary classical performance.
अष्टोत्तरशतकरणानि — सूची च निरन्तरताBharata's fourth chapter does not present its 108 entries as an undifferentiated list. Read across its own textual quartiles, the catalogue moves from grounded, evenly weighted opening configurations toward increasingly asymmetric, aerial, rotational, and compound vocabulary, closing on a return to grounded stability. This Part reads the full catalogue against that internal shape, against the component systems of sthāna, cārī, and nṛtta-hasta that generate it combinatorially, and against the documented, carefully bounded continuity connecting the identified panels of Part Three to a living dancer's own trained body today.
The sthāna inventory from which each karaṇa draws its own single standing base divides, as established in Part One, §2.2, along two independent axes: weight distribution and vertical compression. Samapāda sthānas hold the weight evenly across both legs and dominate the catalogue's opening quartile; ayata and aiḍa sthānas shift the weight asymmetrically onto a single leg and become progressively more frequent through the second and third quartiles; vaiṣṇava and maṇḍala sthānas introduce a deep, wide-based knee-bend supplying the stability more vigorous cārīs require, and cluster disproportionately around the karaṇas immediately preceding the catalogue's own most demanding bhramarī-type entries.
The sixteen cārīs divide into the bhaumī and ākāśikī classes established in Part One, §2.3, and this division correlates directly and legibly with the sculptural record: an ākāśikī-cārī karaṇa raises the working leg above hip height, occupying the upper register of a carved panel's own frame, while a bhaumī-cārī karaṇa keeps both legs within the panel's lower register throughout. Tracking the proportion of ākāśikī to bhaumī cārīs quartile by quartile is, as the tables below demonstrate, the single most direct visual index of the catalogue's own escalating physical intensity.
Where sthāna and cārī together fix the broad physical shape of a karaṇa, it is frequently the nṛtta-hasta component — recakita, ardharecita, and the wider vocabulary of arm-circling and extension terms — that supplies the fine differentiation distinguishing two karaṇas sharing an identical base and leg action, since the nṛtta-hasta inventory is combinatorially larger than the comparatively constrained sthāna and cārī inventories permit on their own, a fact directly relevant to why sculptural identification, per Part Two, §12.2, requires convergence across all three components rather than resting on sthāna or cārī alone.
Beneath these three obligatory components, the recaka layer of Part One, §5.1 supplies a further modifying variable, present or absent, and where present specific to hand, foot, waist, or neck, that can differentiate two karaṇas otherwise identical in sthāna, cārī, and hasta. This four-layer combinatorial system — sthāna, cārī, hasta, and optional recaka — is what allows a comparatively compact set of underlying components to generate the full, non-redundant 108-item catalogue without requiring 108 entirely independent movement types invented from nothing, a structural economy directly parallel to the phoneme-to-word generativity examined in this series' companion treatment of sphoṭa theory.
Each quartile table that follows states, for every tabled entry, the specific sthāna, cārī, and hasta combination the root verse and Abhinavagupta's gloss converge upon, together with the epigraphic sites at which that specific combination has been identified with confidence. Where a component cannot currently be fixed with full confidence — most frequently the specific recaka type on a weathered panel — this register states that uncertainty directly rather than supplying a confident-sounding but unsupported specification.
Read across all four quartiles rather than one at a time, the sthāna inventory shows a clear directional drift within the tabled sample: samapāda and other evenly weighted bases dominate the first-quartile entries, ayata and aiḍa asymmetric bases become the plurality through the second and third quartiles, and vaiṣṇava and maṇḍala deep-stance bases concentrate disproportionately in the fourth quartile, precisely where the ākāśikī and bhramarī vocabulary of §20.2 and Part One, §5.2 demands the greatest lowered-centre-of-gravity stability. This drift is not asserted as a claim about all 108 entries individually — the method-note above governs that caution — but is directly visible even within the smaller, securely tabled sample this register presents, and is consistent with the broader pedagogical-sequencing argument of Part One, §6.3.
The same directional pattern holds, more starkly, for the bhaumī/ākāśikī division: every first-quartile tabled entry is bhaumī or transitional; the second and third quartiles present a roughly even mixture; the fourth quartile is dominated by ākāśikī cārīs to a degree no earlier quartile approaches. This gradient is the single clearest quantitative signature, within the tabled sample, of the catalogue's own escalating physical intensity, and it is the structural fact underlying every quartile-head summary in §21 through §24.
A recurring naming convention worth stating explicitly, because it recurs across multiple quartiles and materially aids identification, pairs a full-form karaṇa name with an ardha- ("half") prefixed counterpart executing a restrained version of the same underlying action — Recitā and Ardharecita in the second quartile, Nikuṭṭaka and Ardhanikuṭṭaka in the third. Where such a pair survives at the same site, as documented for Nikuṭṭaka and Ardhanikuṭṭaka at Chidambaram, the two panels function as mutually confirming evidence, since a secure identification of one member of the pair substantially constrains the plausible identification of the other.
Beyond the full/half pairing convention of §20.8, several named physical motifs recur across widely separated points in the catalogue at progressively greater intensity: the vṛścika (scorpion) motif appears in restrained form in the first quartile (Vṛścikakuṭṭita) and returns intensified in the second (Latāvṛścika); the recaka rotation itself appears in restrained partial form early (Bhramaraka, first quartile) and returns as a complete rotation later (Bhramaraka, full form, third quartile; Ghūrṇita, fourth quartile). This recurring-and-intensifying structure is direct internal textual evidence for the pedagogical escalation argument of Part One, §6.3, operating not merely at the coarse quartile level but within specific named motif-families the compiler appears to have deliberately distributed across the sequence.
Given roughly a dozen commonly attested sthānas, sixteen cārīs, and a nṛtta-hasta inventory considerably larger than either, the theoretical combinatorial space available to the karaṇa system vastly exceeds 108; the closed count is accordingly not a ceiling imposed by combinatorial exhaustion but a selection, from a much larger possible space, of the specific 108 combinations the transmitted tradition treats as canonical — a point that reinforces rather than undercuts the closure argument of Part One, §1.3, since it confirms the number 108 reflects a deliberate act of textual selection rather than a mechanical enumeration of every technically possible combination.
Dominated by samapāda and closely related even-weighted sthānas combined predominantly with bhaumī cārīs, the opening quartile functions pedagogically as well as textually, introducing the fundamental standing bases before the catalogue proceeds to its later, more vigorous vocabulary. Talapuṣpapuṭa, the opening entry, is the single most consistently identified panel across the five-site corpus, and functions as the calibration anchor for the entire identification method.
| No. | Karaṇa | Sthāna / Cārī / Hasta | Epigraphic Correlate |
|---|---|---|---|
| 1 | Talapuṣpapuṭa | Samapāda sthāna; bhaumī cārī, minimal displacement; cupped, blossom-form nṛtta-hasta read as a formal offering gesture. | ChidambaramTanjavur Securely labelled at Chidambaram's eastern gopuram; the corpus's primary calibration entry. |
| 3 | Kaṭicchinnā (early form) | Ayata sthāna with early asymmetric weighting; bhaumī cārī; hasta held close to the torso, prefiguring the pronounced waist-break of the later, fuller Kaṭicchinna (no. 46). | Chidambaram Identified with moderate confidence via commentarial gloss. |
| 10 | Vṛścikakuṭṭita | Maṇḍala-type grounded sthāna; bhaumī cārī with a curled, scorpion-tail leg action; contained hasta pair — the earliest appearance of the vṛścika motif intensified through later quartiles. | Tanjavur Identified on the vimāna's lower tier. |
| 14 | Kuñcita | Samapāda sthāna; bhaumī cārī with a bent, drawn-in (kuñcita) foot placement; recakita hasta. | ChidambaramTiruvannamalai Identified at both sites with matching sthāna reading. |
| 16 | Bhramaraka (early form) | Grounded sthāna; restrained partial kaṭi-recaka; low-intensity introduction of the recaka vocabulary within an otherwise grounded context. | Chidambaram Partial panel condition; identification relies substantially on textual triangulation. |
| 19 | Ūrudhvaṛtaka | Ayata sthāna; bhaumī cārī with the thigh raised to a moderate, sub-hip height, marking a transitional step toward the ākāśikī vocabulary of later quartiles; extended hasta pair. | Tanjavur Identified on the vimāna's lower-mid tier. |
| 22 | Nitamba | Vaiṣṇava sthāna with pronounced hip displacement (nitamba); bhaumī cārī; hasta following the line of the hip. | Chidambaram Identified via convergent commentarial and sculptural evidence. |
| 27 | Vikṣiptākṣiptaka | Ayata sthāna; a thrown-and-drawn-back (kṣipta, doubled) leg action bridging bhaumī and ākāśikī registers; correspondingly extended hasta — the quartile's closing, transitional entry. | ChidambaramTiruvannamalai Positioned at the close of the first gopuram tier at both sites, consistent with its role as a quartile boundary. |
The recaka vocabulary of Part One, §5 becomes the dominant organising principle across this stretch of the catalogue, opening with Ardharecita's own partial rotation and closing with Bhujaṅgāñcitaka's pronounced serpentine torso curve — among the most visually distinctive and least ambiguous configurations in the entire 108.
| No. | Karaṇa | Sthāna / Cārī / Hasta | Epigraphic Correlate |
|---|---|---|---|
| 28 | Ardharecita | Ayata sthāna; bhaumī cārī; a half (ardha-) kaṭi-recaka opening the quartile's own partial-rotation logic. | Chidambaram Clearly labelled; torso torsion legible on weathered sections of the eastern gopuram. |
| 31 | Recitā | Aiḍa sthāna; bhaumī cārī; a full, rather than partial, hasta-recaka distinguishing this entry from the preceding partial forms. | Chidambaram Identified by direct comparison against the adjacent Ardharecita panel. |
| 33 | Latāvṛścika | Vaiṣṇava sthāna; ākāśikī cārī combining a creeper-like (latā) leg extension with the scorpion-curl (vṛścika) motif of the first quartile, now intensified into full aerial extension. | Tanjavur Identified on the vimāna's mid-tier; among the first panels connecting the Chidambaram–Tanjavur comparative sequence. |
| 37 | Dolāpāda | Ayata sthāna; bhaumī cārī with a swinging, pendulum-like (dolā) leg action; hasta following the swing's own arc. | Kumbakonam Identified on the facade's single register. |
| 40 | Daṇḍapakṣa | Ayata sthāna; ākāśikī cārī with the leg extended rod-straight (daṇḍa) at or above hip height; wing-like (pakṣa) spread of the nṛtta-hastas. | ChidambaramKumbakonam Present at both sites with minor stylistic divergence in hasta angle. |
| 43 | Viṣkambha | Maṇḍala sthāna; bhaumī cārī with a broad, propping (viṣkambha) stance; grounded, symmetrically held hasta pair. | Tanjavur Identified with moderate confidence. |
| 46 | Kaṭicchinna | Ayata sthāna; bhaumī cārī; a pronounced lateral break (chinna) at the waist, the quartile's own most legible torsional feature. | Tanjavur Securely identified even at the vimāna's considerable carved height. |
| 50 | Aḍakauśika | Aiḍa sthāna; ākāśikī cārī with a hooked (kauśika) foot placement at raised height; contained hasta. | Chidambaram Identified via convergent commentarial and sculptural evidence. |
| 54 | Bhujaṅgāñcitaka | Vaiṣṇava sthāna; ākāśikī cārī with a bent, drawn-in (āñcita) leg; a full serpentine (bhujaṅga) curve of the torso — the quartile's most visually intense closing entry. | ChidambaramTanjavurTiruvannamalai The most consistently identified karaṇa across three major sites after Talapuṣpapuṭa itself. |
Compound configurations dominate this stretch of the catalogue, combining a daṇḍa-type extended limb, a recaka-type rotation, and a vṛścika-type curl within single karaṇas rather than presenting each independently — a compounding that corresponds, in the sculptural record, to a measurably higher rate of contested identification among researchers than for the first two quartiles.
| No. | Karaṇa | Sthāna / Cārī / Hasta | Epigraphic Correlate |
|---|---|---|---|
| 55 | Daṇḍarecita | Ayata sthāna; ākāśikī cārī with a rod-straight extended limb combined with a full, rather than partial, recaka rotation. | Chidambaram Identification depends on distinguishing full versus partial recaka; Abhinavagupta's gloss is the deciding evidence. |
| 58 | Bhramaraka (full form) | Maṇḍala sthāna; a complete, rather than partial, bhramarī-adjacent rotation building on the restrained first-quartile Bhramaraka (no. 16). | Tanjavur Identified by comparison against the corresponding first-quartile motif. |
| 61 | Nikuṭṭaka | Samapāda sthāna; bhaumī cārī with a stamping, percussive foot-action (kuṭṭ-); contained, close-held hasta pair marking a specific tāla accent. | TanjavurKumbakonam The percussive strike is rendered through a distinctive raised-heel convention shared across both workshops. |
| 65 | Ardhamaṭalī | Aiḍa sthāna; ākāśikī cārī with a partial (ardha-) circling foot-action; recakita hasta. | Chidambaram Identified with moderate confidence. |
| 68 | Ardhanikuṭṭaka | Samapāda sthāna; bhaumī cārī; a restrained, partial counterpart to Nikuṭṭaka's percussive strike (no. 61). | Chidambaram Identified by direct comparison against the adjacent, better-preserved Nikuṭṭaka panel. |
| 72 | Vṛṣastita | Vaiṣṇava sthāna; bhaumī cārī with a wide, bull-stance (vṛṣa) grounded base; hasta held low and symmetrically. | Tanjavur Identified via convergent commentarial and sculptural evidence. |
| 75 | Vṛṣabhakrīḍita | Vaiṣṇava sthāna; a wide, grounded stance with a pawing, percussive leg-action distinct from the aerial ākāśikī cārīs otherwise dominant in this quartile. | Tanjavur One of the more physically distinctive compound karaṇas in the vimāna sequence. |
| 78 | Lalāṭatilaka | Ayata sthāna; ākāśikī cārī with the raised foot brought toward the forehead (lalāṭa) line; hasta framing the leg's own extension. | Chidambaram Identified with moderate confidence; panel condition partial. |
| 81 | Valitaka | Maṇḍala sthāna; a twisting (vali-) rotation of the full torso against a stabilising base — the quartile's own closing pivot into the fourth quartile's bhramarī vocabulary. | ChidambaramTiruvannamalai Positioned at a tier boundary at both sites, consistent with its role as a structural quartile marker. |
Full-body bhramarī rotations and the catalogue's most demanding ākāśikī cārīs dominate this closing stretch, and it is here that the frozen-motion problem of Part Three, §18 becomes most acute, and the compensating sculptural devices described there appear at their highest concentration. The catalogue closes on Vaiśākha's own grounded, symmetrical stability, structurally answering the first quartile's opening Talapuṣpapuṭa.
| No. | Karaṇa | Sthāna / Cārī / Hasta | Epigraphic Correlate |
|---|---|---|---|
| 82 | Ghūrṇita | Maṇḍala sthāna; a full, continuous rotational turn about the vertical axis, distinguished by commentary from the partial rotations of earlier quartiles; hasta held in extension throughout the turn. | Chidambaram Rendered via the exaggerated-torsion convention of Part Three, §18.4. |
| 86 | Recita (fourth-quartile form) | Vaiṣṇava sthāna; ākāśikī cārī with a full hasta-recaka executed at the raised-leg apex. | Tanjavur Identified on the vimāna's upper tier. |
| 89 | Alātaka | Maṇḍala sthāna; a wide, sweeping ākāśikī leg action named for the alāta (firebrand) whose circular swing traces a visible ring when moved quickly. | TanjavurTiruvannamalai The wide leg-sweep is among the most legible ākāśikī renderings in the corpus. |
| 93 | Daṇḍapadmaka | Ayata sthāna; ākāśikī cārī combining the rod-straight (daṇḍa) leg extension with a lotus-form (padma) hasta configuration. | Chidambaram Identified with moderate confidence. |
| 97 | Simhavikrīḍita | Maṇḍala sthāna; a lion's-play (siṃha-vikrīḍita) configuration combining a wide grounded base with a raised, clawing hasta gesture. | Tanjavur Identified via convergent commentarial and sculptural evidence. |
| 101 | [Boundary of the Vṛddhācalam sequence] | The Vṛddhagiriśvara program terminates its own surviving sequence at this point; karaṇas 102–108 are absent from that site specifically. | Vṛddhācalam Final securely surviving panel in the site's own sequence. |
| 103 | Ūrdhvajānu | Aiḍa sthāna; ākāśikī cārī with the knee (jānu) raised well above hip height (ūrdhva); extended hasta pair framing the raised knee. | ChidambaramTanjavur Identified at both major sites with matching sthāna reading. |
| 106 | Ativikrānta | Vaiṣṇava sthāna; an intensified (ati-) striding (vikrānta) leg action nearing the catalogue's own maximum physical extension before the closing resolution. | Tanjavur Identified on the vimāna's uppermost surviving tier. |
| 108 | Vaiśākha | Samapāda sthāna, wide and evenly weighted; bhaumī cārī; hasta held in calm symmetry — the catalogue's own structural resolution, deliberately contrasting the preceding bhramarī-dominated stretch. | ChidambaramTanjavur Positioned at the terminal point of both major sequences; alongside Talapuṣpapuṭa, one of the two most confidently identified panels in the corpus. |
Bharatanatyam's own foundational adavu vocabulary — the basic combinatorial units of araimaṇḍi (half-seated base), foot strikes, and arm positions every student learns before advancing to full compositional repertoire — is widely read by dance scholars as a structurally descended, though not verse-for-verse identical, elaboration of the sthāna–cārī–hasta logic of §20, transmitted through the guru-śiṣya paramparā of the naṭṭuvanār teaching lineages historically associated with several of the temples examined in Part Three, rather than through continuous, verse-by-verse study of the Nāṭyaśāstra's own fourth chapter, which had itself fallen into substantial practical obscurity by the colonial period.
The Nāṭyaśāstra's fourth chapter was restored to active scholarly and pedagogical attention through the twentieth-century revival movements associated with figures including Rukmini Devi Arundale and the Kalakshetra tradition, a restoration that supplied the necessary scholarly precondition — renewed, careful textual attention to a chapter that had lapsed into comparative neglect — for the sculptural identification work examined in Part Two, §12 to be undertaken at all.
Kuchipudi's own historical development, rooted in a distinct Andhra devotional and dance-drama tradition rather than in the Tamil temple-sculptural corpus this register centres on, preserves a related but independently evolved relationship to the same Nāṭyaśāstra source material, sharing Bharata's underlying sthāna and cārī vocabulary at a structural level while having developed its own distinct choreographic conventions, including the tarangam and specific narrative sequences not paralleled in the karaṇa-focused pure-nṛtta tradition this register documents.
It is specifically Padma Subrahmanyam's own twentieth-century reconstructive research program that constitutes the most direct and deliberate act of continuity between the epigraphic record and contemporary performance: her identification of specific temple panels with specific numbered karaṇas was undertaken not as a purely archival exercise but explicitly in order to make those identified karaṇas physically re-performable, and her own subsequent choreographic and pedagogical work, developing what she and her students term Bhāratanṛtyam, represents a documented, traceable case in which stone carved eight to nine centuries earlier was read back into a living dancer's trained body.
The overwhelming majority of contemporary Bharatanatyam and Kuchipudi repertoire is not composed of directly reconstructed, individually identified karaṇas performed in their original Nāṭyaśāstra sequence; it is composed of choreography developed within each tradition's own historically evolved compositional conventions, informed by but not reducible to the karaṇa catalogue examined in this Part. This register states that limit plainly: it does not establish an unbroken, continuous, century-by-century performance tradition of the complete 108 in their original textual sequence.
What the epigraphic and reconstructive record does establish, with considerable confidence, is narrower and more precise: that the 108 karaṇas were once a living, performed vocabulary, textually defined, commentarially glossed, and monumentally inscribed in stone at five specific Tamil temples across several centuries of continuous royal and devotional patronage, and that a documented, scholarly, and pedagogically active bridge now exists, through Subrahmanyam's and subsequent researchers' work, between that historical record and a present-day dancer's own trained body — closing, for these specific identified karaṇas at least, the gap between text, stone, and flesh this entire four-part register has been organised to examine.
The specific identified karaṇas re-performed within the Bhāratanṛtyam tradition are documented and transmitted through the institutional structures Subrahmanyam and her students established specifically for this purpose, a formal pedagogical custodianship distinct from the more diffuse, decentralised transmission characteristic of Bharatanatyam's own adavu vocabulary as discussed in §25.1 — a distinction this register notes because it bears directly on how confidently the reconstructed repertoire's own continued transmission, as opposed to its one-time scholarly recovery, can be projected forward.
Re-performing a karaṇa identified from a weathered stone panel necessarily requires the reconstructing dancer to supply, from her own trained aesthetic judgment, physical detail no static image can specify directly — the precise timing, the specific quality of transition into and out of the identified instant — meaning the reconstructed repertoire, however textually and sculpturally grounded, is not a pure retrieval of an unmediated ancient practice but a scholarly-creative synthesis in which the reconstructor's own contemporary training necessarily participates, a methodological honesty this register considers a strength of Subrahmanyam's own published approach rather than a limitation to be concealed.
Set against comparable text-to-practice reconstruction projects elsewhere in Indological and broader performance-studies scholarship, the karaṇa project's own combination of a closed, numerically fixed textual catalogue, a substantial and geographically distributed sculptural corpus, and a continuously maintained commentarial tradition gives it an unusually strong evidentiary base by the standards of ancient performance reconstruction generally, where textual and material evidence frequently survive independently of each other rather than in the mutually confirming relationship documented across this four-part register.
This four-part register has traced a single object — the 108 karaṇas of the Nāṭyaśāstra's fourth chapter — through four distinct evidentiary registers: the mythological and definitional foundations of Part One, the textual and commentarial transmission of Part Two, the sculptural and architectural record of Part Three, and the catalogue and its living continuity examined here in Part Four. At every stage, this register has attempted to state plainly where the evidence converges with confidence and where it does not, following the evidentiary discipline Subrahmanyam's own triangulated method establishes as the field's own working standard, in the conviction that a scholarly account of a tradition this precisely defined is owed nothing less than the same precision in return.
The Seven-Level Confluence of the 108 Karaṇas with Medical Science, the Natural Sciences, Astro-Science, and Psychology — a single cross-referenced reading of all four preceding modules, mapped against the specific research gaps each domain, taken alone, has left unnoticed.
कायविज्ञानसंगमः — सप्तस्तरीयः सम्बन्धःThe 108 karaṇas have, across the four modules preceding this one, been read as mythology, as text, as stone, and as catalogue. None of those four readings asks a fifth question: what does a system this precisely specified — a closed inventory of coordinated tripartite movement events, transmitted for a millennium through guru-śiṣya kinesthetic pedagogy and monumentalised in a five-site sculptural corpus — look like to a physiologist, a neuroscientist, a psychologist, or a historian of astronomical and calendrical science standing outside the Sanskrit textual tradition entirely? This capstone does not claim the karaṇa system anticipated modern science, nor that modern science has proven what the tradition asserts. It does something narrower and more useful: it identifies, level by level, where a genuine structural correspondence exists, where that correspondence is currently unexamined by either discipline, and where the resulting gap is close enough to the present state of both fields that a specific, fundable, methodologically tractable study could close it.
Reading Part One's tripartite sthāna–cārī–hasta formula and Part Four's quartile catalogue against the vocabulary of kinesiology, biomechanics, and musculoskeletal medicine.
Structural Parallel Clinical biomechanics describes standing posture through a small set of variables — base of support area, centre-of-mass projection relative to that base, and the joint angles maintaining the configuration against gravity. Part One, §2.2's sthāna taxonomy, dividing bases along the same two axes of weight distribution and vertical compression, is formally describing the identical variable set several centuries before the biomechanical vocabulary existed to name it independently. The parallel is structural, not historical: nothing suggests the Nāṭyaśāstra's compilers reasoned from force-plate data, and nothing in modern kinesiology derives from Sanskrit dramaturgy. What the parallel offers is a ready-made, textually fixed, three-thousand-year-stable taxonomy of postural configurations against which a modern biomechanical dataset could, in principle, be directly mapped.
Structural Parallel The sixteen cārīs of Part One, §2.3 each specify a foot position joined to a leg-raising or sweeping action — precisely the unit modern gait analysis calls a kinetic chain, a coordinated sequence of joint rotations propagating from a fixed or moving foot contact. The bhaumī/ākāśikī division maps with unusual cleanliness onto the modern distinction between closed-chain movement (foot maintaining ground contact, force transmitted proximally) and open-chain movement (limb moving freely, force generated distally) — a distinction central to both sports medicine and physical rehabilitation protocol design.
Open Research Gap Musculoskeletal sports-medicine literature on dance injury is disproportionately concentrated on the lower limb — ankle, knee, hip — reflecting where injury incidence is highest in most Western concert-dance forms. The nṛtta-hasta system, by contrast, specifies an equally rigorous and combinatorially rich upper-limb vocabulary carrying no representational content (Part One, §2.4) and therefore no compensatory constraint from narrative legibility, meaning its physical demands are purely biomechanical. No published biomechanical study currently characterises shoulder-girdle and wrist loading across a systematic karaṇa or aṅgahāra sequence. This is a specific, tractable gap: motion-capture and EMG instrumentation of a trained Bhāratanṛtyam performer executing an identified aṅgahāra would generate a dataset with no current equivalent in the dance-medicine literature.
Structural Parallel Part One, §5.1's four recakas — hand, foot, waist, neck — each isolate rotation at a single joint region while the remainder of the body maintains relative stability, a training structure directly analogous to the isolated-joint mobility and motor-control drills used in modern physiotherapy and in dance-specific conditioning programs, where isolating a single rotational degree of freedom before integrating it into compound movement is standard pedagogical practice.
Established The deep knee-bend, wide-base maṇḍala and vaiṣṇava sthānas of Part One, §2.2 place the hip and knee in flexion angles substantially beyond neutral standing posture; orthopaedic and sports-medicine research on deep-squat loading, developed independently for weightlifting and rehabilitation contexts, has established that such positions, correctly executed, increase gluteal and quadriceps recruitment without the elevated shear-force risk once assumed of deep flexion, a finding directly relevant to assessing the loading demands of maṇḍala-based karaṇas even though it was not developed with dance in mind.
Open Research Gap Part Four, §20.6–20.7 documents, within the tabled sample, a directional shift from grounded samapāda-bhaumī combinations in the first quartile toward vaiṣṇava-maṇḍala-ākāśikī combinations in the fourth. No biomechanical study has yet quantified whether this textually documented shift corresponds to a measurable increase in joint torque, ground-reaction force, or metabolic cost across the quartile sequence in actual performance — a study that would test, with instrumentation rather than textual inference alone, whether the catalogue's own internal ordering tracks genuine physical escalation or a purely notional one.
Reading the aṅgahāra sequencing of Part One, §4 and the guru-śiṣya transmission model of Part Two against cerebellar motor sequencing, procedural memory, and the mirror-neuron literature.
Established Cerebellar motor-control research has established that complex movement sequences are stored and executed not as chains of individually retrieved actions but as "chunked" units — a fixed sub-sequence retrieved and executed as a single motor program, with chunk boundaries corresponding to natural break-points in the movement. Structural Parallel The aṅgahāra's own definitional requirement that its constituent karaṇas be executed in fixed, non-negotiable order (Part One, §4.3) describes, independently and centuries earlier, the identical chunking structure cerebellar research would later formalise — a fixed sequence of sub-units whose reordering constitutes an execution error rather than a stylistic variant, precisely the criterion cerebellar chunk-boundary research uses to identify a genuine motor chunk.
Established Procedural memory — motor skill retained and executed without requiring conscious step-by-step recall — is well documented in cognitive neuroscience as functionally and neuroanatomically distinct from declarative memory, engaging basal ganglia and cerebellar circuits relatively independent of hippocampal declarative systems. Structural Parallel Part Two, §9.2's argument that the karaṇa root verses function as retrieval cues for knowledge already held in a trained body, rather than as self-sufficient written instruction, is describing precisely the procedural-memory relationship modern cognitive neuroscience formalises: the verse cues procedural retrieval, it does not encode the movement declaratively.
Established The mirror-neuron system — cortical motor neurons that activate both when an action is performed and when the same action is observed in another — is well documented as a candidate neural substrate for observational motor learning generally. Open Research Gap No published study has examined mirror-system activation specifically during guru-directed karaṇa correction, a pedagogical context distinguished from generic observational learning by the guru's own simultaneous verbal cueing (the sūtra retrieval-cue mechanism of 2.2) alongside direct physical correction — a combined verbal-observational-tactile transmission modality that current mirror-neuron research, focused predominantly on purely observational paradigms, has not yet modelled.
Established The basal ganglia are established as central to the temporal gating of rhythmic motor sequences, with documented involvement in both rhythm perception and the precise timing of sequential motor output. Open Research Gap Because karaṇa execution is tāla-correlated — the rhythmic-cycle framework Part Two, §10.3 identifies as under-specified in Bharata's own root text relative to Śārṅgadeva's later elaboration — a trained performer's basal ganglia timing mechanisms are being driven by an externally imposed rhythmic cycle of unusual metrical complexity relative to Western rhythmic paradigms typically used in basal-ganglia timing research; no study has yet used tāla-cycle karaṇa performance as a naturalistic probe of basal ganglia timing precision under high metrical complexity.
Open Research Gap Part Four, §25.4 documents a small, identifiable population of dancers trained specifically in Subrahmanyam's own reconstructed Bhāratanṛtyam repertoire, executing individually identified karaṇas rather than the broader, less textually anchored adavu vocabulary. This population is neuroscientifically distinctive: their motor training is anchored to a closed, textually fixed 108-item catalogue rather than to an open, continuously evolving repertoire, offering a rare opportunity for functional neuroimaging (fMRI or EEG) of expert motor sequencing against a fixed rather than an open-ended skill set — a study design current dance-neuroscience research, focused overwhelmingly on Western concert-dance forms with open repertoires, has not yet had access to.
Reading the bhramarī vocabulary of Part One, §5.2 and the fourth-quartile catalogue of Part Four, §24 against vestibular physiology, spatial-orientation research, and balance rehabilitation.
Established Sustained rotation about the body's vertical axis drives the vestibulo-ocular reflex, the semicircular-canal-mediated mechanism stabilising gaze during head rotation, and produces post-rotational nystagmus and subjective vertigo in untrained individuals whose habituation threshold has not been raised through repeated exposure. Established Professional dancers and figure skaters who train sustained spinning are independently documented in vestibular research as exhibiting measurably reduced post-rotational nystagmus and subjective vertigo relative to untrained controls, attributed to central vestibular habituation rather than to any change in peripheral semicircular-canal sensitivity. Structural Parallel The bhramarī-dominant fourth quartile's own placement at the catalogue's most physically demanding closing stretch (Part Four, §24) is structurally consistent with a deliberately progressive vestibular-habituation training curriculum, whether or not the tradition's own compilers possessed any concept of the vestibular system as such.
Established Trained rotational performers across multiple dance traditions independently employ a gaze-fixation technique — briefly re-fixing the eyes on a stationary visual reference point during each rotation before snapping the head through the turn — documented in vestibular research to reduce the visual-vestibular mismatch otherwise driving disorientation during sustained spinning. Open Research Gap No published source, textual or ethnographic, currently documents whether an equivalent gaze-fixation technique is specified or implicitly required for bhramarī-type karaṇa execution, despite the movement's own rotational demands being comparable in magnitude to the pirouette and chaîné turns for which the technique is well documented in Western dance pedagogy — a specific, answerable question for ethnochoreological fieldwork with currently practising Bhāratanṛtyam performers.
Established Single-leg stance with the contralateral limb raised well above hip height, the physical configuration the ākāśikī cārī class specifies (Part One, §2.3), is established in balance-research literature as substantially increasing postural-control demand relative to bilateral stance, requiring rapid proprioceptive and vestibular integration to maintain equilibrium — the same physiological demand underlying single-leg-stance balance testing as a standard clinical assessment tool for fall-risk in geriatric and neurological populations.
Open Research Gap Vestibular rehabilitation for patients with chronic dizziness or post-concussion vestibular dysfunction relies on graded exposure protocols whose progression curves are typically derived from generic physiotherapy literature rather than from any dance-specific training tradition; the karaṇa catalogue's own documented progression from grounded first-quartile stability toward fourth-quartile bhramarī intensity (Part Four, §20.6–20.7) constitutes an independently developed, centuries-tested graded-exposure curriculum whose structure has not yet been examined by vestibular rehabilitation researchers as a potential model system, despite its close structural resemblance to the graded-exposure principle rehabilitation protocols already rely on.
Open Research Gap Because karaṇa training under the naṭṭuvanār lineage system (Part Four, §25.1) proceeds through a documented, multi-year pedagogical progression from grounded adavu-level work toward advanced rotational repertoire, it offers a naturalistic, longitudinal population in which vestibular habituation could be tracked against a known training curriculum — a study design distinct from, and potentially more informative than, the cross-sectional expert-versus-novice comparisons dominating current dance-vestibular research.
Reading the nṛtta/nṛtya/nāṭya trichotomy of Part One, §4.1 against embodied cognition, procedural-declarative memory dissociation, flow-state research, and the clinical field of dance movement therapy.
Established Embodied-cognition research in cognitive psychology has established that motor and perceptual systems participate directly in cognitive processing rather than serving merely as input/output peripherals to an amodal central processor — meaning purely formal, non-representational movement is not cognitively inert but engages genuine cognitive and affective processing distinct from, though not necessarily lesser than, the processing engaged by representational or narrative movement. Structural Parallel The karaṇa system's own definitional insistence that nṛtta carries no narrative or referential content (Part One, §4.1, §2.4) positions it, within an embodied-cognition framework, as a purposefully constructed test case for movement's own cognitive and aesthetic value independent of meaning-carrying content — a theoretical position embodied-cognition researchers have arrived at independently, without reference to this or any other specific dance tradition.
Established Kinesthetic or motor memory is well documented as functionally dissociable from declarative memory, with patients exhibiting severe declarative amnesia nonetheless retaining and improving upon previously trained motor skills — clear clinical evidence that procedural motor knowledge is stored and retrieved through mechanisms independent of the explicit, narratable memory the same patients cannot access. Structural Parallel This dissociation directly substantiates, from an independent clinical-psychological literature, the sūtra-compression argument of Part Two, §9.2: a terse root verse can function as an adequate transmission vehicle precisely because it is addressing procedural rather than declarative memory systems in an already-trained reader.
Established Dance movement therapy is an established clinical discipline, recognised by relevant professional accrediting bodies, using structured movement sequences to address a range of psychological and psychosomatic presentations, with a developing evidence base particularly in mood regulation, body-image disturbance, and trauma-related dissociation. Open Research Gap No published dance movement therapy protocol currently incorporates karaṇa-derived movement sequences specifically, despite the system's own combination of precise structural definition, graded physical progression (Level One, §1.6), and formally non-narrative content — a combination that, on the profile dance movement therapy research uses to select or design intervention sequences, would make the karaṇa catalogue a plausible candidate source, currently unexamined.
Established Flow-state research in psychology has established a well-replicated profile of optimal engagement — clear goals, immediate feedback, and a close match between perceived challenge and skill — associated with reduced self-referential processing and heightened absorption, documented across domains including music performance and athletic competition. Structural Parallel The karaṇa system's own tightly specified, closed-catalogue structure, its immediate tāla-cued feedback, and its graded quartile-based challenge progression (Level One, §1.6) map closely onto the flow-state precondition profile, suggesting karaṇa performance is a strong candidate context for flow-state induction, though no study has yet tested trained performers against standard flow-state psychometric instruments during karaṇa-specific execution as distinct from general dance performance.
Open Research Gap Because karaṇa-adjacent training is transmitted through at least two documented distinct pedagogical routes — the diffuse adavu-level Bharatanatyam lineage and the more formally bounded Bhāratanṛtyam reconstruction lineage (Part Four, §25.7) — a comparative psychometric study of attentional control, body-awareness measures, and flow-state disposition across performers trained predominantly within each lineage would test whether the more textually and structurally bounded reconstruction lineage produces measurably different cognitive-psychological profiles than the more diffusely transmitted adavu tradition, a specific comparative design no published study has yet undertaken.
Reading the tāla correlation of Part Two, §10.3 and the temple liturgical calendar of Part Three, §7.4 against circadian and ultradian rhythm research, breath-movement coupling, and autonomic regulation.
Established Ultradian rhythms — biological cycles shorter than twenty-four hours, including respiratory rate, heart-rate variability cycling, and attentional oscillation — are established as entrainable by external periodic stimuli, a phenomenon well documented in the music-cognition literature on rhythmic entrainment. Structural Parallel A sustained tāla cycle, driving the timed execution of a full aṅgahāra sequence, constitutes a specific, high-precision instance of exactly this externally imposed rhythmic entrainment, positioning trained karaṇa performance as a naturalistic model system for ultradian entrainment research that current music-cognition studies, typically using simpler and shorter rhythmic stimuli, have not directly examined.
Established Circadian rhythm research has established that physical performance capacity, including flexibility, reaction time, and cardiovascular readiness, varies systematically across the twenty-four-hour cycle, with most populations showing measurable performance peaks in the late afternoon. Open Research Gap Part Three, §7.4's documentation of ritual tāṇḍava performance tied to Chidambaram's own liturgical calendar raises a specific, currently unexamined question: whether the temple's own traditional performance timings correlate with, precede, or are indifferent to the circadian performance-optimum window independently established in chronobiology — a question answerable through liturgical-calendar analysis cross-referenced against standard circadian performance research, requiring no new physiological data collection.
Established Coupling of respiratory rhythm to movement rhythm is documented in exercise physiology as influencing autonomic balance, with paced, movement-synchronised breathing associated with measurable shifts toward parasympathetic dominance relative to unpaced breathing during equivalent physical exertion. Open Research Gap No published physiological study has measured heart-rate variability or respiratory coupling during sustained aṅgahāra-length karaṇa sequences specifically, despite the tāla-cued, tightly timed structure of such sequences making them an unusually well-controlled stimulus for testing breath-movement coupling relative to the more variably paced movement typically used in autonomic-regulation exercise research.
Open Research Gap Combining the biomechanical loading question of Level One, §1.6 with the autonomic-regulation question of 5.3 above, a single wearable-sensor study tracking heart-rate variability across a performer's progression through all four textual quartiles in a single extended session would test, with physiological rather than purely textual evidence, whether the catalogue's own documented directional shift toward more demanding vaiṣṇava-maṇḍala-ākāśikī combinations (Part Four, §20.6–20.7) produces a correspondingly measurable autonomic load progression — a study requiring only standard, already-available heart-rate-variability instrumentation and a single cooperating trained performer.
Reading the closure argument of Part One, §1.3, the pañcakṛtya schema of Part One, §1.4, and the gopuram orientation evidence of Part Three against the history of Indian astronomical and calendrical science.
Established The number 108 recurs across multiple independently documented domains of classical Indian astronomical and cosmological reckoning — including specific traditional approximations relating solar and lunar diameters to their respective orbital distances, and its use as a standard count in astronomically timed ritual recitation cycles — a numerological convergence documented in the history of Indian astronomical science independently of the Nāṭyaśāstra. Structural Parallel The karaṇa catalogue's own closure at exactly 108 (Part One, §1.3) is consistent with, though not thereby proven to be causally derived from, this broader numerological convention; this register does not assert a specific astronomical derivation for the dance catalogue's own count, since no surviving textual source makes that derivation explicit, but notes the convergence as worth stating precisely because Part One, §1.3 already flags the number's broader numerological resonance without settling its origin.
Established Classical Indian temple architecture is documented, across a substantial body of vāstu-śāstra literature and archaeoastronomical fieldwork at specific sites, as frequently incorporating deliberate astronomical orientation, including alignment to solstice sunrise axes and to specific nakṣatra risings relevant to a temple's own consecration date. Open Research Gap No published archaeoastronomical survey has specifically examined whether the eastern gopuram's own karaṇa-panel sequence at Chidambaram (Part Three, §13.1), or the equivalent sequences at the other four sites, show any correlation between panel position and solar or nakṣatra alignment at the point of construction — a survey requiring only standard archaeoastronomical surveying methods already applied to these temples' other architectural features.
Established Classical Indian cosmological literature independently documents an elaborated system of nested cosmic time cycles (yuga, kalpa, and their further subdivisions), a chronological framework developed within Purāṇic and astronomical literature broadly contemporaneous with, though textually distinct from, the Nāṭyaśāstra's own compilation. Structural Parallel Part One, §1.4's pañcakṛtya schema — creation, preservation, destruction, concealment, and grace, read at Chidambaram as continuous with the karaṇa catalogue's own kinetic completeness — shares its own cyclical, creation-to-dissolution logical structure with the cosmic time-cycle framework, a shared cyclical logic this register notes as a genuine formal resonance within the broader Śaiva cosmological universe both belong to, without claiming one system was derived from the other.
Open Research Gap Where a temple's own consecration date can be fixed with confidence through the paleographic dating method of Part Two, §7.5, a specific and currently untested question follows directly: does the karaṇa panel sequence's own degree of fidelity to Bharata's textual order (tracked site by site in Part Three, §11 across the original register) correlate with any astronomically significant date in that temple's own consecration calendar, or is panel-sequence fidelity determined purely by the architectural and workshop factors already documented in Part Three — a question answerable by cross-referencing two datasets, paleographic dating and panel-sequence fidelity, that already exist independently in the scholarly record but have not yet been directly compared against each other.
What no single one of the preceding six levels can see on its own, and a ranked agenda for closing the gaps each level has identified individually, ordered by proximity to the current state of each discipline.
Read individually, Levels One through Six each isolate a single disciplinary lens — biomechanics, neurology, vestibular physiology, psychology, chronobiology, astro-science — applied to a system the tradition itself never divided along those lines. Part One, §2.1's own insistence that a karaṇa is one indivisible event, not three sequential components, applies with equal force across the disciplinary division this capstone has just imposed for expository purposes: a single executed karaṇa is simultaneously a biomechanical loading event, a cerebellar motor-chunk retrieval, a vestibular-proprioceptive integration, a psychologically absorbing formal experience, and a tāla-entrained autonomic event, all within the same single count of performed time. No individual discipline studies this integration as such, because no individual discipline's own methods are designed to capture more than one of these five simultaneous dimensions at once.
Biomechanics alone cannot explain why the fourth-quartile bhramarī vocabulary is pedagogically sequenced last rather than distributed randomly (that requires the vestibular-habituation and motor-chunking evidence of Levels Two and Three together). Vestibular physiology alone cannot explain why the training curriculum is embedded in a closed, numerically fixed catalogue rather than an open-ended progressive drill set (that requires the closure argument of Level Six read against the psychological absorption evidence of Level Four). Psychology alone cannot explain why the system's aesthetic theory insists so precisely on non-referential form (that requires the biomechanical and neurological specificity of Levels One and Two, which only a genuinely non-referential formal system could be compressed into sūtra verses terse enough to function as the procedural-memory retrieval cues Level Two, §2.2 describes). Each level's own blind spot is filled only by at least one other level, which is the central empirical argument for treating the karaṇa system as a genuinely integrative case study rather than parcelling it out to whichever single discipline happens to notice it first.
Nothing in Levels One through Six should be read as claiming the Nāṭyaśāstra's compilers possessed anticipatory knowledge of cerebellar chunking, vestibular habituation curves, or ultradian entrainment mechanisms; the claim throughout has been structural correspondence, independently arrived at by two traditions of inquiry separated by a millennium and by entirely different methods of verification — textual-commentarial transmission on one side, controlled experimental replication on the other. Where this capstone has tagged a claim Structural Parallel, that tag is doing real interpretive work: it is the difference between a legitimate cross-disciplinary reading and a fabricated ancient-science claim, and it is not offered as a hedge but as the single most important methodological commitment this entire document makes.
Ranked by how directly each open gap can be closed with instrumentation, personnel, and methods that already exist — requiring no new theory, only new application — the following table orders the eleven open research gaps identified across Levels One through Six from nearest to furthest from immediate feasibility.
| Rank | Gap | Level | Nearest-Point Rationale |
|---|---|---|---|
| 1 | Consecration-date vs. panel-sequence-fidelity cross-reference | 6.4 | Both datasets already exist in the published scholarly record; requires comparison, not new fieldwork. |
| 2 | Liturgical-calendar vs. circadian-performance-window correlation | 5.2 | Requires only calendar analysis against already-established chronobiology; no new physiological data needed. |
| 3 | Archaeoastronomical survey of gopuram panel-position alignment | 6.2 | Uses standard, already-applied archaeoastronomical surveying methods on structures already under active study. |
| 4 | Heart-rate variability across a single quartile-progression session | 5.4 | Requires one cooperating performer and standard, inexpensive wearable HRV instrumentation. |
| 5 | Ethnochoreological documentation of bhramarī gaze technique | 3.2 | Fieldwork with currently practising performers; no laboratory equipment required. |
| 6 | Motion-capture and EMG of upper-limb loading in an aṅgahāra sequence | 1.3 | Standard biomechanics-lab equipment; requires only a trained performer and lab access. |
| 7 | Comparative psychometric profiling across transmission lineages | 4.5 | Standard psychometric instruments; requires recruiting performers from two identifiable lineages. |
| 8 | Vestibular habituation tracked across the training timeline | 3.5 | Longitudinal design; requires sustained access to a training institution over multiple years. |
| 9 | Flow-state psychometrics during karaṇa-specific performance | 4.4 | Standard flow-state instruments; requires isolating karaṇa-specific from general dance-performance flow. |
| 10 | fMRI/EEG of reconstructed-repertoire experts against fixed catalogue | 2.5 | Requires neuroimaging access and a small, specific expert population; higher cost and recruitment barrier. |
| 11 | Biomechanical loading comparison across all four textual quartiles | 1.6 | Requires full-catalogue motion-capture across an extended multi-session protocol; the most resource-intensive item on this agenda. |
What this capstone has attempted is neither an assertion that ancient India possessed a covert science of vestibular rehabilitation or cerebellar chunking, nor a dismissal of the karaṇa system as merely mythological material irrelevant to a modern researcher's own concerns. It is a specific, disciplined claim: that a system this precisely specified, this thoroughly documented across text, commentary, and stone, and this centrally organised around the trained human body's own capacities, constitutes a genuinely under-examined natural object for at least six distinct scientific literatures, each of which has, until now, had no occasion to look at it. The eleven gaps tabled in 7.4 are not speculative in the sense of being unanswerable; they are speculative only in the narrower sense that no one has yet asked the question with the instruments already sitting in the relevant laboratories. Closing them would not prove the karaṇa tradition was secretly scientific. It would do something more modest and more valuable: it would tell us, with actual data rather than inference from eight-hundred-year-old stone alone, exactly how a body organised by this specific tradition works.