| Metadata | Value |
|---|---|
| Status | Advisory — informs modelling; ragamalika not yet re-extracted |
| Version | 1.0.0 |
| Last Updated | 2026-09-10 |
| Author | carnatic-musicologist review (for Seshadri) |
| Document Type | Musicology review |
| Backs | database/migrations/V47__demerge_ragamalika_visvanatham_from_natabharanam.sql |
Muthuswami Dikshitar. Reviews how this composition should be modelled after migration V47 separated the ragamalika lyric variants that had been incorrectly merged into the Natabhranam krithi during import. V47 is applied; the ragamalika still needs re-extraction as its own krithi, and the recommendations below describe how it should be structured when that happens.
Current Database State The ragamalika composition does not yet exist as its own krithi row. What exists today is:
Krithi ab99b150: titled “viSva nAthaM bhajEhaM”, stored as musical_form = KRITHI, is_ragamalika = false, primary_raga = Natabhranam, with a single krithi_ragas row (Natabhranam, order_index 0). It has 2 sections: PALLAVI and SAMASHTI_CHARANAM. The lyric text is the Natabhranam-raga krithi only (the non-ragamalika composition). The pending migration V47__demerge_ragamalika_visvanatham_from_natabharanam.sql correctly identifies that ragamalika lyric variants were incorrectly merged into this Natabhranam krithi during import. It removes those variants and resets the import sources for re-extraction as a separate krithi. This is the correct approach.
Once re-extracted, the ragamalika will need its own krithi row. Below are my recommendations for how to model it.
Answers to Your Questions
order_index section_type label 1 PALLAVI Pallavi 2 ANUPALLAVI Anupallavi 3 CHARANAM Charanam The raga subsections within each structural section are not separate krithi_sections. They are internal subdivisions of the sahitya text within each section. The structural form is Pallavi / Anupallavi / Charanam – the standard Krithi form. The raga changes happen within each section, and that raga-to-text mapping is captured by the krithi_ragas junction table (with section indicating which structural section each raga belongs to).
Rationale: The domain model (section 6.2) says to represent the raga sequence via KrithiRaga rows with orderIndex and optional section. The section-type enum does not have a “ragamalika subsection” type, nor should it – these are not structural divisions in the way Pallavi/Anupallavi/Charanam are. The raga transitions are a property of the ragamalika form overlaid on the standard section structure.
This is a Krithi by form (Pallavi-Anupallavi-Charanam structure, sahitya-dominant, no Muktayi Swaram or Ettugada pattern). The ragamalika aspect is an additional property captured by the flag and the krithi_ragas rows.
Tala Confirmed: Adi tala (Chatusra-jati Triputa, 8 beats). This is well-attested in all published sources for this composition.
Return Ragas (Viloma) Classification The viloma (return) ragas at the end of the Charanam section – going back through Bhupala to Sama, then Mohana through Sri – belong to the Charanam section in the krithi_ragas.section column.
Musically, the viloma is a structural feature of this composition where the composer revisits the ragas in reverse order. But the performance practice is clear: after the forward Charanam ragas (7-14), the singer continues in the Charanam without returning to the Anupallavi or Pallavi as separate structural restarts. The entire viloma passage is sung as a continuation of the Charanam. The Anupallavi viloma (6 to 1) similarly stays within the Anupallavi section.
So the section assignment should be:
Anupallavi viloma ragas (Mohana back to Sri): section = ‘anupallavi’ Charanam viloma ragas (Bhupala back to Sama, then Mohana back to Sri): section = ‘charanam’
order_index raga section 1 Sri pallavi 2 Arabhi pallavi 3 Gauri anupallavi 4 Nata anupallavi 5 Gaula anupallavi 6 Mohana anupallavi 7 Mohana (viloma) anupallavi 8 Gaula (viloma) anupallavi 9 Nata (viloma) anupallavi 10 Gauri (viloma) anupallavi 11 Arabhi (viloma) anupallavi 12 Sri (viloma) anupallavi 13 Sama charanam 14 Lalita charanam 15 Bhairava charanam 16 Saranga charanam 17 Sankarabharanam charanam 18 Kambhoji charanam 19 Devakriya charanam 20 Bhupala charanam 21 Bhupala (viloma) charanam 22 Devakriya (viloma) charanam 23 Sankarabharanam (viloma) charanam 24 Kambhoji (viloma) charanam 25 Saranga (viloma) charanam 26 Bhairava (viloma) charanam 27 Lalita (viloma) charanam 28 Sama (viloma) charanam 29 Mohana (viloma) charanam 30 Gaula (viloma) charanam 31 Nata (viloma) charanam 32 Gauri (viloma) charanam 33 Arabhi (viloma) charanam 34 Sri (viloma) charanam Important modeling note: The viloma entries reference the same raga_id as their forward counterparts (e.g., order_index 6 and 7 both point to the same Mohana raga row). The notes column on krithi_ragas should carry “viloma” to distinguish the return passage. The order_index captures the performance sequence unambiguously.
For primary_raga_id on the krithi row: since the domain model says “a single primaryRagaId is insufficient for a ragamalika,” you could set it to Sri (the opening raga) as a convenience/display default, or leave it NULL. I recommend Sri as the primary since the composition begins and ends in Sri ragam.
Standard ragamalika: Each section has its own raga(s), proceeding forward only. Example: Dikshitar’s “Kamalamba Navavarana” cycle (each krithi in one raga, no viloma). Simple ragamalika with viloma: Some compositions reverse a short sequence at the end to “close the circle.” Example: a 5-raga malika that returns 5-4-3-2-1 at the end. The “viSva nAthaM bhajEhaM” is exceptional because:
It uses 14 ragas (caturdaSa ragamalika) – already rare. It applies viloma within each major section (not just at the very end). The Anupallavi reverses through all ragas heard so far (6 to 1), and the Charanam reverses through all 14 ragas (14 to 1). The total number of raga transitions in performance is therefore 34, not 14. This palindromic/arch structure within sections is a compositional tour de force specific to this work. It is not the norm for Dikshitar’s other ragamalikas (such as the Navagraha or Panchalinga krithis, which are sets of individual compositions rather than multi-raga single compositions). Among single-composition ragamalikas in the broader tradition, this level of viloma elaboration is rare and marks the piece as a showcase of compositional mastery.
Findings Summary Incorrect is_ragamalika = false on the existing krithi row ab99b150. However, this row is actually the Natabhranam krithi (the non-ragamalika composition with the same title), so it is correctly marked false. The ragamalika simply does not exist as a separate krithi yet. Ragamalika lyric variants merged into wrong krithi. The V47 migration addresses this correctly by removing the 6 ragamalika lyric variants from the Natabhranam krithi and resetting import records for re-extraction. Suspect (recommend human review at re-extraction time) Charanam viloma scope: The guru-guha blogspot source shows the final viloma in the Charanam going all the way back to Sri (raga 1), passing through the Anupallavi and Pallavi ragas. Some published editions treat the final six ragas (Mohana through Sri) as a return to the Anupallavi/Pallavi sections rather than a continuation of the Charanam. I have assigned them to charanam above based on performance practice (the singer does not structurally restart the Anupallavi/Pallavi), but this deserves verification against the Semmangudi or T.K. Govinda Rao recordings when the composition is re-ingested. Primary raga assignment: When the new ragamalika krithi row is created, confirm whether primary_raga_id should be Sri (opening raga) or NULL. The domain model discourages collapsing to a single raga but does not explicitly forbid a “lead” raga designation for display purposes. Cosmetic Raga name normalization: Ensure the raga entries in the database use consistent transliteration: “Sankarabharanam” (not “Shankarabharanam”), “Bhupala” (not “Bhoopalam”), “Devakriya” (not “Deva Kriya” with a space). Check that each of the 14 ragas has a corresponding row in the ragas table before creating krithi_ragas entries. The V47 migration is the correct first step. After it runs and re-extraction produces the ragamalika as a new krithi, the ingestion pipeline should create the row with the structure described above: 3 sections, 34 krithi_ragas rows with viloma notes, is_ragamalika = true, and musical_form = ‘KRITHI’.