01 Mythological Origins of the Karaṇas

The Nāṭyaśāstra frames the one hundred and eight karaṇas not as a human choreographic invention but as a revelation descending through a specific transmission chain: Śiva, in his form as Naṭarāja-Maheśvara, is said to have performed the tāṇḍava at the close of the fourth chapter's dramatic frame, and to have instructed his attendant Taṇḍu in the grammar of this movement. Taṇḍu in turn transmits the science — hence the alternate designation of the corpus as tāṇḍava lakṣaṇa, "the defining marks of Taṇḍu's dance" — to Bharata, who codifies it within the received text of the Nāṭyaśāstra.

तेन ताण्डवसंज्ञं तु प्रोक्तं नाट्यं मयानघ । तांडुनोक्तं यथान्यायं तांडवं तेन तत्स्मृतम् ॥ Nāṭyaśāstra 4.13 (recension-dependent numbering) — the tāṇḍava is named for Taṇḍu, through whom Bharata received its codification.

The closing verse of the karaṇa-list section makes explicit the soteriological claim attached to correct performance: one who dances the karaṇas well, "created by Maheśvara," is said to be freed from sin and to attain the abode of the deity. This is not incidental praise-poetry; it fixes the karaṇas within a devotional-liturgical frame rather than a purely aesthetic one, which is precisely why their sculptural placement is concentrated on temple gopurams — thresholds through which the devotee physically passes — rather than in secular performance halls.

Read against the broader vāk-doctrine architecture of this series, the Śiva–Taṇḍu–Bharata chain can be understood as a structural parallel to the descent of sound through parā, paśyantī, madhyamā and vaikharī: an undifferentiated, ecstatic divine movement (comparable to parā) is progressively particularized through Taṇḍu's mediating discipline (paśyantī/madhyamā) into the codified, teachable, repeatable karaṇa-units of vaikharī — spoken śloka and enacted gesture converging as two vaikharī registers of a single descending revelation. This parallel is offered as an interpretive frame for this series and is not attested in the Nāṭyaśāstra or its commentaries.

Within the Sastras Extended corpus's larger comparative method, the Tandu-transmission narrative also invites comparison with other guru-mediated revelation chains recorded elsewhere in the sastric tradition, such as the descent of grammatical insight through Panini's own reported divine hearing, or the transmission of Sri Vidya ritual sequencing discussed in Part Six of this series. In each case a single, ultimately superhuman source is mediated through exactly one named intermediary before reaching the human codifier whose name the resulting text bears. Whether this single-intermediary pattern reflects a genuine shared convention of sastric legitimation across otherwise unrelated domains, or is simply how any origin-narrative tends to compress a longer and less tidy historical transmission process, remains an open question this module does not resolve.

02 Definition and Anatomy of a Karaṇa

Bharata's working definition binds three elements into a single indivisible unit of dance: hasta (hand position, specifically the nṛtta-hasta or pure-dance hand gesture, as opposed to the meaning-bearing hastas of aṅgikābhinaya), pāda (the footwork, realized through a cārī), and sthāna (the standing posture or stance of the torso and lower body). The formula circulating in secondary literature — hasta + pāda + sthāna = karaṇa — compresses the root text's own phrasing.

हस्तपादसमायोगो नृत्तस्य करणं भवेत् Attributed to Nāṭyaśāstra 4, cited widely in secondary dance-scholarship as the operative definition of karaṇa; exact chapter-verse placement varies by manuscript recension and should be checked against a specific critical edition before citation in print.

Abhinavagupta's gloss in the Abhinavabhāratī sharpens this further by specifying that gati (gait/movement of the feet) and sthiti (the settled stance) must be samīlita — "harmoniously combined" — with the nṛtta-hasta, such that a karaṇa is never a static tableau but always a brief, complete arc of motion: an entry, a held or transitional configuration, and an implied continuation into the next karaṇa within a sequence.

Padma Subrahmanyam's philological and kinesthetic re-reading insisted on precisely this point against a long twentieth-century habit of treating karaṇas as static "poses": she argued, on the basis of both the verbal root kṛ ("to do/act") underlying the noun karaṇa and the sculptural evidence of transitional limb-blur in several temple panels, that each unit encodes a movement-phrase, not a frozen shape, a position she developed at length in Karanas: Common Dance Codes of India and Indonesia.

A further point worth flagging for the practicing reconstructor: because the karana definition names three components, or four in Abhinavagupta's elaboration, without specifying their relative timing within the movement-phrase, two readings equally faithful to the Sanskrit stanza can still disagree sharply on tempo and sequencing, that is, whether the hasta configuration is established before, during, or after the footwork's cari is completed. This ambiguity is almost certainly the single largest source of the divergence documented in subsection 25 between competing modern reconstruction lineages, since a printed stanza cannot specify timing in the way a demonstrated movement or a notation system with explicit temporal markers could.

03 The Concept of Mārgī versus Deśī

The Nāṭyaśāstra situates the karaṇa corpus within the category of mārgī — the "path" of pan-Indian, translocal, classically codified performance, understood as capable of spiritually elevating the spectator — as distinct from deśī, regionally inflected, vernacular or entertainment-oriented performance whose remit is enjoyment rather than transcendence. The karaṇas, along with the aṅgahāras built from them, are explicitly assigned to the mārgī register: they are the grammar of a classical, transregional dance-language rather than a folk repertoire tied to a single locality.

This distinction carries direct art-historical weight. Van Kooij and other scholars of the Tamil temple reliefs have noted that the near-identical sequencing of karaṇa panels across geographically separated sites — Chidambaram in the Kaveri delta, Thiruvannamalai further south, Vriddhachalam inland — testifies to the operation of a shared mārgī template circulating among sthapatis (temple architect-sculptors) and possibly transmitted through the Nāṭyaśāstra text itself or an intermediary iconographic manual, rather than to independent local invention at each site.

The mārgī/deśī polarity also offers a useful lens for later Bharatanatyam and Kuchipudi reconstruction debates (see subsection 25): where a modern guru's adavu vocabulary diverges from the karaṇa descriptions, the divergence can often be read as a deśī accretion — regionally transmitted through guru-śiṣya-paramparā — layered atop, or partially substituting for, a mārgī substrate that had itself grown fragmentary by the medieval and early-modern periods. This is a heuristic proposed here, not a documented historical process.

The Natyasastra's own framing ties the margi designation not merely to geographic reach but to ritual efficacy: margi performance is described as capable of conferring merit and spiritual benefit on both performer and audience in a way desi performance, oriented purely toward entertainment, is not. This is the textual root of the temple gopuram placements documented in Part III; the karanas are carved at thresholds precisely because their margi status makes them appropriate content for a space mediating between the ordinary world and the sanctum.

04 The Role of Nṛtta, Aṅgahāras and Maṇḍalas

Nṛtta is pure, non-representational dance — movement organized by rhythm and geometry rather than by narrative or emotional signification (which belongs instead to nṛtya and abhinaya proper). The karaṇa is the atomic unit of nṛtta. Karaṇas combine, in the Nāṭyaśāstra's own architecture, into aṅgahāras — "limb-garlands," sequences most commonly of six or seven karaṇas strung together into a longer phrase — and aṅgahāras in turn combine into even longer units. A maṇḍala denotes a further class of complex, often circular or multi-directional movement-combinations, particularly implicated in the more acrobatic and combative choreography described in later chapters of the text and elaborated at length in Abhinavagupta's eleventh-chapter commentary on maṇḍala-vikalpana.

करणान्यङ्गहाराश्च रेचकाश्चैव ते त्रयः । नृत्तस्य हि शरीरं तु त्रिविधं परिकीर्तितम् ॥ Paraphrased structural summary of Nāṭyaśāstra 4's own architecture: karaṇas, aṅgahāras and recakas together constitute the threefold body of nṛtta. Exact verse numbering should be verified against a specific critical edition.

Sreenivasa Rao's summary of Abhinavabhāratī notes that aerial (ākāśa) maṇḍalas are traditionally enumerated in ten varieties and ground-based (bhaumaṇḍala) maṇḍalas in eight, a classification useful for reading the more athletic karaṇa-clusters (e.g., the Vṛścika group, subsection 23) which combine multiple caris across an implied maṇḍala-path rather than a single static footprint.

It is worth noting that the Natyasastra's own three-tier architecture of karana, angahara, and the still-larger combinations built from angaharas mirrors a compositional logic found across many Sanskrit sastric systems, where an atomic unit such as a phoneme, karana, or sutra combines into an intermediate unit such as a word, angahara, or adhikarana, which in turn combines into a complete utterance or performance. Whether this compositional parallel between the grammar of language and the grammar of dance reflects a shared underlying theory of composition in the intellectual culture that produced both, or is simply a natural way for any complex human skill to be taught in graduated stages, is a question this register raises without claiming to answer.

05 Essential Movement Categories: Rechita, Bhramara, Maṇḍala

Three recurring movement-qualities structure a large proportion of karaṇa names and their kinetic content. Rechita denotes a whirling, releasing or "emptying-out" rotation of a limb — visible in the sheer density of karaṇa names built on this root (Svastikarechita, Ardharechita, Ākṣiptarechita, Vaiśākharechita, Daṇḍarechita, Vṛścikarechita, and others across the 108). Bhramara names a circling, bee-like orbital motion, most explicitly in karaṇa 38, Bhramaraka. Maṇḍala-type movement, as noted in subsection 4, describes a compound circuit combining several cārīs, most visible in karaṇa 53, Cakramaṇḍala, where the entire body describes a turning circle.

Because rechita-type karaṇas cluster disproportionately in the 28–54 and 82–108 registers of the traditional numbering (subsections 22 and 24 below), a plausible structural reading — offered here as hypothesis rather than documented Nāṭyaśāstra commentary — is that the text's compilers grouped the corpus not strictly by narrative order of Śiva's dance but partly by kinetic family, interleaving quieter sthāna-dominant units with more kinetically extreme rechita and vṛścika (scorpion-type) units to create a danceable dynamic contour across a performed aṅgahāra sequence.

A further observation: the disproportionate concentration of rechita-type names in the 28 to 54 and 82 to 108 ranges noted above roughly brackets the corpus's two most kinetically extreme thematic clusters, the Vrscika or scorpion family in the second block and the Simha or lion and related animal-sport family in the fourth. This suggests that rechita as a movement-quality may function less as an independent kinetic category and more as a connective or transitional technique used to link the corpus's calmer sthana-dominant units to its more athletically demanding named postures, a reading that remains a working hypothesis rather than an attested structural principle.

06 The Significance of the Fourth Chapter: Tāṇḍava Lakṣaṇa

Chapter Four of the Nāṭyaśāstra, titled Tāṇḍava Lakṣaṇa ("the defining characteristics of tāṇḍava"), is structurally unusual within the text: it opens not with technical definition but with a narrative frame recounting Śiva's dance and its transmission, before pivoting — traditionally around stanzas enumerating the 108 names in sequence, followed by individual descriptive stanzas for each — into pure technical codification. This dual character, myth-frame plus technical appendix, is part of why the chapter has attracted disproportionate epigraphic and sculptural attention relative to other technical chapters of the text: it supplies both the theological warrant (why this dance matters, who performed it, what merit accrues) and the operational content (what, precisely, must be carved or danced) in a single unit.

Kapila Vatsyayan's technical re-interpretation of the chapter, published through the Sangeet Natak Akademi and later digitized by the Indian Culture portal, remains a primary secondary-source anchor for reading the chapter's terse descriptive stanzas back into embodied movement, and is cited throughout Parts III and IV of this register alongside Padma Subrahmanyam's independent reconstruction.

The chapter's dual mythic-technical structure also explains a recurring feature of the secondary scholarship surveyed throughout this module: authors working primarily as dance-reconstructionists such as Subrahmanyam tend to foreground the technical descriptive stanzas, while authors working primarily as art historians or epigraphists, such as van Kooij and the Thiruvadigai field survey, tend to foreground the mythic frame and its sculptural placement. This module has attempted to hold both readings together rather than privileging either, consistent with the three-tier evidentiary separation stated at the outset.

07 The Role of Grantha Script in Temple Inscriptions

The karaṇa panels at the four gopurams of Chidambaram carry, at the east and west gateways specifically, accompanying inscriptions in Grantha script — the script historically used across the Tamil country for recording Sanskrit texts, as distinct from Tamil script proper — giving the relevant Nāṭyaśāstra verse-name directly beneath or beside the corresponding sculptural panel. This co-presence of Grantha-inscribed śloka and carved image is the single most important piece of evidence establishing, beyond stylistic inference, that the reliefs are a deliberate and identified representation of Bharata's specific 108-unit corpus rather than a generic dance-frieze drawing loosely on dance iconography.

The choice of Grantha — a script optimized for Sanskrit phonology, retaining distinctions (aspirated consonants, vowel-length markers) that Tamil script does not cleanly support — over the vernacular script otherwise used in the temple's administrative and donative inscriptions is itself a statement: the karaṇa program is being presented as mārgī, pan-Indian, scripturally anchored text made visible, not as a locally composed devotional caption. This reading is consistent with, but not independently attested beyond, the mārgī/deśī distinction discussed in subsection 3.

It is also worth noting what the Grantha inscriptions do not do: they identify the karana by name and citation but do not, in the documentation surveyed for this module, include the full descriptive stanza carved alongside the image. This suggests the inscriptions functioned as an identifying label for a devotee or visiting scholar already familiar with the fuller textual tradition, rather than as a self-contained teaching aid for someone encountering the karana corpus for the first time at the temple itself.

08 Deciphering Abhinavabhāratī for Movement Interpretation

Abhinavagupta's Abhinavabhāratī, composed roughly in the late tenth to early eleventh century, remains the earliest surviving full commentary on the Nāṭyaśāstra and the primary lens through which the terse, often single-line karaṇa descriptions of the root text are expanded into fuller kinesthetic instruction. For Chapter Four specifically, Abhinavagupta supplies the compact formula distinguishing gati (gait), sthānaka (stance), cārī (leg-position) and nṛtta-hasta (dance hand-gesture) as the four components whose sammilana (harmonious union) constitutes karaṇa — a gloss that both confirms and slightly extends Bharata's own three-part formula (subsection 2) by separating gati from sthāna as analytically distinct even though co-present in performance.

Multiple published translations and critical editions of the Abhinavabhāratī's fourth-chapter portion exist (K. Krishnamoorthy's volume one, and M. Ramakrishna Kavi's volumes two and three, among the most cited), and reconstruction scholars including Subrahmanyam treat convergence between these translations and the surviving sculptural evidence as the strongest available corroboration for any specific karaṇa reading, precisely because the commentary and the temple reliefs derive from largely independent transmission streams separated by both centuries and geography.

Because Abhinavagupta wrote roughly two to three centuries before the earliest of the datable Chidambaram gopuram reliefs, per Appendix F, his commentary cannot have been directly informed by that specific sculptural program, which strengthens rather than weakens its evidentiary value. Where a modern reconstructor finds Abhinavabharati's textual gloss and the independently dated Chidambaram relief converging on the same reading of a karana's stance or hand-position, the convergence is genuinely cross-checking two separate transmission streams rather than one informing the other.

09 The Use of Short Pithy Verses for Movement Instructions

Each of the 108 karaṇas receives, in the root Nāṭyaśāstra text, a single terse stanza — rarely more than two lines — specifying stance, foot-action and hand-position in a compressed, sūtra-like register closer to technical shorthand than to descriptive poetry. This economy of language is functional rather than merely stylistic: it mirrors the mnemonic compression found throughout Sanskrit śāstric literature generally (compare the sūtra style of Pāṇini's Aṣṭādhyāyī within this series' own vāk-doctrine cluster), designed for oral retention by a practicing dancer-in-training rather than for silent reading.

पुष्पपुटौ तु करौ वामपार्श्वे प्रयोजयेत् । अग्रतलसंचारः पादश्च सन्नतः स्मृतः ॥ Descriptive stanza for karaṇa 1, Talapuṣpapuṭa, paraphrasing the received gloss: the hands are held in puṣpapuṭa (cupped-flower) form on the left side, the foot moves in agratalasañcāra, and the side (torso) is held sannata (bent/inclined). Wording follows the widely circulated secondary paraphrase; the reader should consult a critical edition for the exact root-text reading before scholarly citation.

The instruction is skeletal by design — it names the sthāna, the cārī and the hasta by their already-technical names (defined earlier in the text, in the chapters on hastas and on cārīs) rather than re-describing them in plain language, meaning the karaṇa-stanza functions as a kind of index or pointer into the fuller technical vocabulary developed across the preceding chapters.

The terse, index-like character of the descriptive stanzas also has a direct pedagogical implication worth stating plainly: a stanza of this kind could only ever have functioned as a memory aid for a student who had already received the full embodied technique directly from a teacher, not as a stand-alone instruction manual from which an unguided reader could learn a karana from cold. This is part of why the sculptural record in Part III carries such disproportionate evidentiary weight in modern reconstruction; it is, in a real sense, the only surviving trace of the embodied demonstration the stanzas originally presupposed.

10 Comparative Study: Abhinaya Darpaṇa and Saṅgīta Ratnākara

Nandikeśvara's Abhinaya Darpaṇa, a considerably later and shorter manual widely used in present-day Bharatanatyam pedagogy, does not reproduce the karaṇa corpus in full but does preserve overlapping technical vocabulary — sthānas, cārīs, and a reduced set of hand gestures — allowing points of cross-verification for individual technical terms even where the Darpaṇa's own performance vocabulary (built around the later adavu system) diverges structurally from karaṇa sequencing. Śārṅgadeva's thirteenth-century Saṅgīta Ratnākara, particularly its Nartanādhyāya (chapter on dance), engages the Nāṭyaśāstra's dance material directly and in several places offers its own restatement or gloss of karaṇa-adjacent categories, making it — alongside Abhinavabhāratī — one of the few pre-modern texts through which the karaṇa tradition can be triangulated across centuries.

Where the Ratnākara's Nartanādhyāya terminology diverges from Bharata's own (for instance in expanded cārī or sthāna counts), this register treats the divergence as evidence of a living, regionally differentiated performance tradition continuing to elaborate the Nāṭyaśāstra's base grammar across the intervening centuries, rather than as evidence that either text is in error — a stance consistent with this series' general refusal to flatten historical textual variation into false synthetic agreement.

Because the Abhinaya Darpana's adavu-oriented vocabulary postdates the karana corpus by several centuries and was composed for a substantially different pedagogical context, namely training a performer in a reduced, theatrically focused technique rather than the full margi corpus, this module treats agreement between the Darpana and a specific karana reading as suggestive corroboration only, never as decisive confirmation on its own, a caution that applies equally to the Ratnakara's Nartanadhyaya material.

11 Structural Mapping of the 108 Kinetic Units

The received sequence of 108 karaṇa names is fixed in the root text's own enumerative stanzas (traditionally located around Nāṭyaśāstra 4.33–44 in commonly cited editions, though numbering varies by recension), after which each name receives its individual descriptive stanza in the same order. This fixed ordinal sequence — not alphabetical, not thematically grouped in any stated way — is the backbone against which every subsequent sculptural program (Chidambaram, Sarangapani, Thiruvannamalai, and the partial programs at Thanjavur and Vriddhachalam) is compared, since panel numbering and panel identification depend on matching the carved sequence to this textual order.

Modern tabulations (Vatsyayan, Subrahmanyam, and the epigraphic surveys underlying the asianart.com and related documentation of the Thiruvadigai and Chidambaram panels) generally number the corpus 1–108 following this received order, and this register adopts the same numbering throughout Part IV for direct comparability with the published scholarship cited in subsections 13–17.

One consequence of the karanas' fixed, non-thematic ordinal sequence is that any thematic grouping proposed by this module in Appendix B, or by any other secondary source, is necessarily an external analytic overlay rather than a feature of the transmitted text itself, and should be labelled as such whenever it is used, a labelling convention this module has tried to observe consistently through its evidentiary tags.

12 Modern Textual Reconstructions: Padma Subrahmanyam's Methodology

Padma Subrahmanyam's twentieth-century reconstruction project remains the single most influential modern intervention in karaṇa studies. Working simultaneously from the Sanskrit descriptive stanzas, Abhinavagupta's commentary, and firsthand fieldwork across the sculptural programs at Chidambaram, Thanjavur, Kumbakonam, Thiruvannamalai and Vriddhachalam — and later extending her comparative fieldwork to the reliefs at Prambanan in Indonesia — she argued for treating each karaṇa as a complete movement-phrase rather than a static pose, a methodological stance that produced her own performance style, subsequently distinguished from mainstream Bharatanatyam under the name Bharatanṛtyam precisely because several established gurus (including figures associated with the Kalakshetra tradition and prominent Kuchipudi lineages) reconstructed the same textual stanzas into visibly different embodied movement, with disagreement extending down to the level of individual cārī identification.

This persistent scholarly disagreement is itself significant evidence for this register's argument (developed across Part III) that no single "correct" reconstruction can be verified from textual description alone; the sculptural record functions not merely as illustration of an already-settled reading but as an independent, and in places the only decisive, evidentiary check on what a given karaṇa-stanza could plausibly have specified in embodied practice.

The scale of divergence among Subrahmanyam's, the Kalakshetra tradition's, and the Kuchipudi reconstructions of the same stanzas is sometimes treated in popular writing as evidence that one lineage must simply be mistaken and another correct. This module takes a more cautious position: absent a surviving continuous performance transmission unambiguously tracing back to the Natyasastra period itself, the available evidence supports treating each major reconstruction as a serious, textually and sculpturally grounded hypothesis rather than as either a definitively solved or definitively mistaken reading.

13 Chidambaram Nataraja Temple: The Complete 108-Karaṇa Program

Chidambaram is the only one of the five canonically studied sites to preserve what scholarship treats as a genuinely complete representation of all 108 karaṇas, distributed across the passages of all four gopurams of the temple complex. The east and west gopuram panels are accompanied by Grantha-script citation of the corresponding Nāṭyaśāstra verse (subsection 7), which is the primary reason Chidambaram functions as the anchor site against which the more fragmentary programs at the other four temples are calibrated. The series across all four gopurams is largely — though not perfectly — identical in content and sequencing; the east, south and west gopurams are generally dated to the twelfth and thirteenth centuries, with the north gopuram somewhat later.

Because Chidambaram is Naṭarāja's own principal shrine, the placement of the complete karaṇa cycle at all four cardinal gopurams — rather than concentrated at a single entrance — can plausibly be read as a deliberate spatial claim: the entire built perimeter of the temple becomes, in effect, a four-fold repetition of Śiva's dance of creation, so that a devotee circumambulating the town-temple complex physically re-enacts the mythological transmission narrated in subsection 1 at each cardinal threshold. This spatial-theological reading is proposed here and is not attested as an explicit statement in any inscription at the site.

The near-identical repetition of the karana cycle across all four of Chidambaram's gopurams, despite their differing construction dates per Appendix F, also suggests that later builders working on the north gopuram deliberately matched the iconographic program already established at the three earlier gateways rather than introducing variation, evidence of a strong site-specific convention that later patrons and sthapatis chose to preserve rather than revise.

14 Brihadisvara Temple, Thanjavur: The Hidden Vimāna Reliefs

The Bṛhadīśvara (Rājarājeśvaram) temple at Thanjavur, completed under Rājarāja I Chola in the early eleventh century, carries its own karaṇa program, but in a markedly different spatial register from Chidambaram's public gopuram passages: the Thanjavur karaṇa reliefs are placed high on the vimāna (the temple's principal tower over the sanctum), in a location not designed or intended for close public viewing at ground level and only properly legible via scaffolding, telephoto photography, or, in the modern era, drone and high-resolution survey documentation. This placement — sacred, elevated, and functionally hidden from ordinary devotional sightlines — sets Thanjavur apart from the deliberately public, processional Chidambaram program.

The contrast between Chidambaram's public gopuram placement and Thanjavur's elevated, near-inaccessible vimāna placement suggests two distinct theological logics for the same sculptural content: one oriented toward the devotee's physical passage and repeated visual encounter (Chidambaram), the other oriented toward offering to the deity or to a purely cosmological viewership rather than to the human worshipper (Thanjavur) — an interpretive distinction offered here as hypothesis rather than as an attested doctrinal statement from either temple's own inscriptional record.

It is worth asking why Rajaraja I's building campaign, otherwise famous for its monumental public inscriptions recording endowments and administrative detail in accessible locations, chose to place its karana program in the temple's least publicly visible location. One plausible reading, offered here as hypothesis, is that the vimana reliefs were conceived as an offering directed toward the enshrined deity rather than toward the human worshipper, consistent with the broader Chola-period practice of reserving certain iconographic content for the sanctum's immediate vicinity rather than for public gopuram passages.

15 Sarangapani Temple, Kumbakonam: External Façade Reliefs

The east gopuram of the Sarangapani temple at Kumbakonam — notable as a Vaiṣṇava temple carrying a Śaiva-liturgical dance program, itself a datum worth pausing on — depicts a more-or-less complete karaṇa series performed, unusually among the five sites, by a male dancer figure rather than the female dancer-with-musicians configuration found at Chidambaram. The reliefs are placed on the external façade of the gopuram rather than within an interior passage, making them directly visible from the temple's exterior approach.

The presence of a substantially complete Nāṭyaśāstra karaṇa cycle on a Vaiṣṇava temple gopuram is consistent with this series' broader argument (developed elsewhere in the Śāstras Extended corpus) that mārgī performance-grammar in this period functioned as a shared pan-sectarian technical language rather than as exclusively Śaiva liturgical property, even where the mythological frame-narrative of the karaṇas (subsection 1) is explicitly Śaiva. This reading is offered as a working hypothesis for the series and not as an established art-historical consensus.

The Sarangapani program's placement on the external facade, fully visible from outside the temple precinct rather than confined to an interior passage, represents a third distinct spatial logic beyond Chidambaram's processional-interior and Thanjavur's elevated-hidden models, that is, a karana cycle displayed, in effect, to the wider town rather than only to those who physically enter the temple, a placement choice that would repay closer comparison with other externally faced sculptural programs of the same period.

16 Arunachalesvara Temple, Thiruvannamalai: Systematic Raja Gopuram Arrangement

The sixteenth-century Rāja Gopuram of the Arunachalesvara temple at Thiruvannamalai carries its karaṇa program across twenty pilasters, each bearing nine vertically arranged panels, for a total of one hundred and eighty individual dance-figure panels, of which one hundred and eight are identified by scholarship as corresponding specifically to the Nāṭyaśāstra karaṇa series, the remainder comprising supplementary dance and mythological figures outside the strict 108-unit corpus. This numerically systematic pilaster arrangement — a fixed grid of panels per pilaster, repeated across a fixed number of pilasters — represents the most architecturally regularized of the five sites' karaṇa programs.

The 180-panel/108-karaṇa ratio at Thiruvannamalai, if the identification is correct, implies that roughly forty percent of the gopuram's dance-figure panels were reserved for material outside the strict Bharata corpus — plausibly deśī or locally devised supplementary dance imagery coexisting with the mārgī karaṇa core on the same architectural surface, a coexistence that would itself illustrate, rather than contradict, the mārgī/deśī distinction discussed in subsection 3. This numerical inference is offered as a plausible reading of the published panel counts and should be checked against a dedicated panel-by-panel identification study before being treated as established.

The sheer scale of the Thiruvannamalai program, twenty pilasters bearing nine panels each, also makes it, among the five sites, the one whose supplementary non-karana dance imagery is most extensively documented alongside the core 108-unit cycle, offering a comparatively rich dataset for future study of how desi or locally devised dance imagery was visually integrated with margi karana content on a single monumental surface.

17 Vriddhagirishvara Temple, Vriddhachalam: The Incomplete 101-Karaṇa Layout

Related field documentation of a comparable inland Chola-period temple at Thiruvadigai — a four-armed Shiva dancing the first eighty-one karaṇas in a horizontally arranged relief on the right-hand side of a clockwise-followed passage, with the series left incomplete and located in a passage not originally intended for public viewing — provides a directly comparable case-type for the kind of partial, non-completed karaṇa program traditionally reported at Vriddhachalam's Vṛddhagirīśvara temple, where scholarship records an incomplete layout of roughly 101 of the 108 units rather than the full corpus found at Chidambaram.

Incomplete karaṇa programs of this kind — whether Vriddhachalam's reported 101-unit layout or Thiruvadigai's 81-unit sequence — raise an open question this register does not attempt to resolve with false certainty: whether incompleteness reflects an unfinished building campaign interrupted by war, patronage collapse or dynastic transition (a mundane historical explanation), or a deliberate iconographic choice to represent only a portion of Śiva's dance at a site of secondary liturgical importance relative to Chidambaram. Both explanations remain live possibilities in the cited literature, and this register flags the question as genuinely open rather than adjudicating it.

Comparative documentation of incomplete karana programs at both Vriddhachalam and Thiruvadigai, if the two sites' incompleteness turns out on closer study to share a common cause, would strengthen the case for reading incomplete programs as a recognized category of secondary-shrine iconography in this period rather than as isolated accidents of individual building-campaign history, though, as stated in the main text above, this module treats the question as genuinely open rather than resolved.

18 "Frozen Motion" versus Real-Time Kinetic Action in Sculpture

A recurring methodological tension in karaṇa iconography concerns how a medium capable only of representing a single static instant — stone relief — can be read as evidence for what the Nāṭyaśāstra itself insists (subsection 2) is fundamentally a movement-phrase rather than a pose. Reconstruction scholarship generally resolves this by treating the carved image as capturing the single most diagnostically legible instant within the implied arc of motion — typically a moment of maximal extension, torsion or balance — from which a trained reconstructor works backward and forward to infer the fuller kinetic phrase, rather than treating the carved instant as the entirety of the prescribed movement. Several panels across the five sites show visible attempts at conveying motion within the static medium itself: overlapping or doubled limb outlines in a small number of panels, drapery and ornament rendered as if still in flight, and asymmetric weight distribution that would be structurally unstable as a held pose but is entirely legible as a captured mid-transition instant.

This "frozen motion" problem is not unique to karaṇa iconography — it recurs across the broader history of dance sculpture globally — but it carries particular weight here because the sculptural evidence is being used not merely to illustrate an already-understood text but, in a number of contested cases, as primary evidence for resolving genuine ambiguity in the terse Sanskrit descriptive stanzas (subsection 9) themselves, meaning methodological caution about over-reading a single frozen instant as a complete movement-specification is directly load-bearing for the field's larger reconstruction claims.

The frozen-motion problem also bears directly on how confidently this module's own Part IV panel-identification notes should be read: an identification marked tentative or contested in the cited literature often reflects precisely this difficulty, a carved instant compatible with more than one candidate karana's implied arc of motion, rather than a simple failure of prior scholarship to look carefully enough at the panel in question.

19 The Iconography of Shiva and Parvati Figures

Across the five sites, the dancer figure executing the karaṇa sequence is variously rendered: at Chidambaram, predominantly a female dancer accompanied by two musicians per panel; at Sarangapani, a male dancer figure (subsection 15); and in the Thiruvadigai comparanda, a four-armed Shiva figure explicitly performing the sequence in propria persona rather than through a human devadasi-proxy. This variation matters because it bears directly on how each site's program relates to the karaṇas' own Śiva–Taṇḍu–Bharata mythological frame (subsection 1): where Shiva himself is the carved dancer, the panel functions as direct cosmological depiction; where a human female dancer performs the sequence (as at Chidambaram's gopurams), the panel instead depicts the temple's devadasi tradition — women dedicated to temple service who performed the sixteen traditional forms of ritual hospitality to the deity, dance among them — enacting Shiva's own dance as devotional service, a subtly different representational claim.

Parvati's presence in the broader iconographic program of these gopurams (though not, in the cited documentation, typically as the performer of the karaṇa sequence herself) is generally as witness-figure within the larger Naṭarāja narrative rather than as co-dancer of the 108-unit corpus specifically; where Pārvatī does appear dancing in South Indian sculpture it is more often in connection with the distinct lāsya tradition than with the tāṇḍava karaṇa corpus proper, a distinction this register flags to avoid conflating two related but textually separate performance categories.

The devadasi-performer convention documented at Chidambaram also connects this module's art-historical material to the broader institutional history of temple dance service in South India, a history whose modern discontinuation, the devadasi system was formally abolished by twentieth-century legislation, forms part of the historical backdrop against which Subrahmanyam's and other twentieth-century reconstruction projects discussed in subsection 12 should be understood: the reconstructors were not merely recovering lost technique from stone and text, but doing so in the wake of the institutional tradition that had, for centuries, been the karanas' primary living transmission context.

20 Static Postures (Sthāna), Leg Movements (Cārī), Arm Gestures (Nṛtta Hastas)

Before the individual karaṇa catalogue, the Nāṭyaśāstra separately codifies its three constituent vocabularies. Sthānas — standing postures — are organized under named categories including Vaiṣṇava, Samapāda, Vaiśākha, Maṇḍala, Ālīḍha and Pratyālīḍha, roughly forty in total across six broad groupings by scholarly count, each with specified applications to particular dramatic or dance contexts. Cārīs — leg and foot movements — are divided into bhaumī (ground-level) and ākāśikī (aerial) categories. Nṛtta hastas are the pure-dance hand configurations, distinguished from the meaning-bearing hastas used in narrative abhinaya even though several hasta-shapes are shared vocabulary between the two systems.

The four Nyāyas (Bhārata, Sāttvata, Vārṣagaṇya and Kaiśika) referenced in the broader Nāṭyaśāstra apparatus for regulating weapon-handling and combative staged movement intersect with the maṇḍala-vikalpana material (subsection 4) at points relevant to the more physically extreme karaṇas of the 55–108 range (Parts IV.3–IV.4 below), particularly the Vṛścika (scorpion) and Siṃha (lion) groups, whose named postures draw on the same combative-movement vocabulary.

Every karaṇa entry below states, where the received descriptive stanza specifies it, its sthāna, cārī and nṛtta-hasta component and its literal name-meaning; where a specific temple panel identification has been published in the cited secondary literature, that identification is noted with its Modern Scholarship tag and treated as tentative unless the source itself claims certainty, since several panels remain contested between two or more candidate karaṇa identifications in the cited surveys.

The separation of sthana, cari and nrtta-hasta into independently catalogued vocabularies before the karana list itself is proper is pedagogically significant: it implies that a Natyasastra-trained dancer was expected to master each of the three component systems on its own terms first, only afterward learning how specific combinations of already-known stances, foot-movements and hand-gestures are named and packaged as the 108 karanas. This layered pedagogy, component vocabulary before compound unit, mirrors the layered pedagogy of Sanskrit grammar itself, where individual dhatus and pratyayas are learned before the sutras that govern their combination, a structural parallel this series has noted elsewhere in its treatment of vaikhari and codified gesture.

21 Postures 1 to 27: Talapuṣpapuṭa to Urdhvajānu

The opening twenty-seven karaṇas establish the corpus's foundational vocabulary of svastika (crossed-limb) configurations, nikuṭṭaka (shouldering) actions, and rechita (whirling) releases, several of which recur as compounding elements in later, more complex units.

Within this opening block, the svastika-family names (7, 8, 13, 15, 16, 17, 22) form the single densest thematic cluster of the entire 108-unit corpus, suggesting that crossed-limb configuration functioned as the corpus's primary organizing grammar for its foundational teaching sequence, with more elaborate rechita, vṛścika and siṃha vocabulary (Parts IV.2–IV.4) introduced only once this base grammar is established — a pedagogical-structural reading offered as hypothesis, not as an explicit statement in the root text.

Beyond the svastika-family clustering already noted, this opening block also introduces, in Bhujangatrasita (24), the corpus's first serpent-referencing name, anticipating the much larger serpent and scorpion family that dominates the corpus's kinetically extreme middle and later sections. Its early placement, well before the dense Vrscika cluster of the second block, may function as a controlled first exposure to serpentine imagery before the more demanding scorpion-family postures are introduced, though this reading remains speculative.

22 Postures 28 to 54: Ardhamattallī to Uromaṇḍala

The second block introduces the corpus's most kinetically extreme sub-family, the Vṛścika (scorpion) group (42, 44, 46, 47), alongside continued rechita elaboration and the first appearance of animal-referencing names that recur throughout the remainder of the corpus.

This second block's Vṛścika cluster (42, 44, 46, 47) is the corpus's first sustained run of combative, ālīḍha/pratyālīḍha-associated postures, and its placement roughly at the numerical midpoint of the full 108-unit sequence is consistent with the pedagogical-structural reading proposed in subsection 21: a foundational block of svastika and rechita vocabulary is mastered first, after which the more physically demanding scorpion-family units are introduced once the dancer's base technique is secure. The block also closes on Cakramaṇḍala (53) and Uromaṇḍala (54), both explicitly maṇḍala-associated units, which may function as a structural cadence returning the sequence to the circling, whole-body movement established as a base category in subsection 4 before the corpus proceeds into its third block.

23 Postures 55 to 81: Ākṣipta to Sarpita

This block moves through swinging and whirlpool-type actions (Dolāpāda, Āvarta) into the needle-family (Sūcī, Ardhasūcī, Sūcividdha) and closes with animal-referencing grace figures (Mayūralalita, Sarpita).

That the Thiruvadigai relief sequence, wherever it is treated as terminating at karaṇa 81 (Sarpita), stops precisely at the close of this third block rather than at an arbitrary panel-count is suggestive: it may indicate that the sthapatis working at Thiruvadigai treated karaṇas 1–81 as a coherent sub-unit suitable for a shorter devotional program, distinct from the full 108-unit cycle reserved for Chidambaram's principal shrine. This remains speculative and is not stated in the cited field documentation.

The needle-family units in this block, Sucividdha, Ardhasuci and Suci themselves, together with the earlier Gandasuci (71), represent the corpus's most precise and least expansively kinetic register, standing in deliberate contrast to the swinging, whirlpool-type openings of the same block, Dolapada and Avarta. Read as a sequence, the block moves from broad, sweeping actions toward increasingly pointed, small-amplitude gestures before closing on the animal-referencing grace of Mayuralalita and Sarpita, a contour that, if intentional, would model a classic performance dynamic of expansion followed by contraction and refinement.

24 Postures 82 to 108: Daṇḍapāda to Gaṅgāvataraṇa

The final block closes the corpus with a dense sequence of animal-referencing names (Hariṇapluta, Karihasta, Siṃhavikrīḍitaka, Siṃhākarṣita, Elakākrīḍita, Vṛṣabhakrīḍita, Nāgāpasarpita) before culminating in karaṇa 108, Gaṅgāvataraṇa — the descent of the Ganga — which closes the sequence on an explicitly cosmogonic image tied to Śiva's own iconography as the one who receives the descending river in his matted locks.

The corpus's closure on Gaṅgāvataraṇa rather than on a numerically neater or more kinetically climactic unit reinforces the reading proposed in subsection 1: the karaṇa sequence is bookended by cosmogonic imagery — implicit creation-through-dance at its mythological opening, explicit reception of the descending, world-purifying Gaṅgā at its close — framing the entire technical corpus of pure nṛtta within an encompassing cosmological narrative rather than presenting it as a neutral, purely technical catalogue of movement.

The concentration of animal-referencing names in this final block, deer, elephant, lion twice, ram, bull, serpent, is denser than in any other block of the corpus, and together with the block's cosmogonic close on Gangavatarana suggests that the compilers reserved the corpus's most vivid, mythologically resonant imagery for its final movement, building toward the culminating image of Siva receiving the descending Ganga rather than distributing such imagery evenly across the full 108 units.

25 Continuity of Karaṇas into Contemporary Classical Dance

Direct, unbroken transmission of the full 108-karaṇa corpus into any single living guru-śiṣya-paramparā is not documented; scholarship generally agrees that by the time of the major twentieth-century reconstruction efforts, only a fragmentary subset of karaṇas and their derivative movement-vocabulary had survived in active practice within Bharatanatyam, Kuchipudi, Odissi and related traditions, embedded largely unrecognized within the adavu systems taught by name rather than by explicit karaṇa-reference. Padma Subrahmanyam's Bharatanṛtyam repertoire (subsection 12), Adyar Lakshman's Kalakshetra-tradition reconstructions, and Kuchipudi reconstructions associated with Vempati Chinna Satyam and C. R. Acharya each represent independent twentieth-century attempts to rebuild practiced technique from the textual and sculptural record, and — as noted in subsection 12 — these independent reconstructions diverge from one another significantly, in places down to disagreement over which cārī a given karaṇa specifies.

This pattern — a codified classical corpus surviving in fragmentary, name-disconnected form within regional practice before being consciously reconstructed against the root text and its sculptural record in the modern period — recurs across several other domains treated elsewhere in the Śāstras Extended corpus (compare the fate of the vaikharī-level grammatical apparatus described in Module One's Extension Argument, and the fragmentary survival of Śrī Vidyā ritual sequencing discussed in Part Six of this series). Whether this recurring pattern reflects a genuine structural regularity in how codified Sanskrit performance-knowledge transmits across the medieval-to-modern rupture, or is simply the visible shape of what survives any sufficiently long transmission chain regardless of content, is a question this register raises but does not claim to settle.

It is also worth registering, without overstating the claim, that the very existence of multiple independent twentieth-century reconstruction projects, rather than a single uncontested revival, is itself evidence of the karana corpus's enduring cultural authority: the Natyasastra's fourth chapter remained sufficiently prestigious across the modern classical dance revival period that multiple serious practitioners judged it worth the very considerable philological and kinesthetic labour of independent reconstruction, rather than treating the corpus as a purely antiquarian curiosity.