Lightning above the six wheels, and the name at last

Taḍit is lightning and latā is a creeper, the climbing vine that winds up a tree; the compound sets them in apposition, lightning that is a vine, and the point of comparison is shape as much as brilliance. A lightning stroke is thin, vertical and branching, and so is the ascending power the previous verses described. The image is not this hymn's invention. It is the standard figure for kuṇḍalinī across the literature, and the Saundaryalaharī uses the same word in the same context, calling her a slender streak of lightning stationed above the six lotuses. Readers comparing editions will find the first syllable spelled either taḍit or taṭit, with retroflex ḍ or retroflex ṭ; both are attested, the variation is old, and nothing turns on it.

Ṣaṭcakroparisaṃsthitā is the hymn's own statement of how the centres are counted. Six wheels, and she is above them — not the seventh in a series but past the series altogether. This is the ordinary reckoning in the tradition, and it explains a small puzzle in modern popular accounts, which usually list seven cakras and thereby flatten a distinction the sources cared about. The six are enumerable, each with its element, its petal count and its syllable; what is above them is not another item of the same kind. The verb in saṃsthitā carries the prefix sam, which intensifies: not merely placed but settled, established, standing firm.

The third name is the one where editions part company. The text as printed here reads mahāsaktiḥ with a dental s, and admits two resolutions — mahā plus āsakti, or mahā plus sakti, both from √sañj, to cling or adhere — either of which gives a great attachment. A great many editions and reciters instead have mahāśaktiḥ with a palatal ś, the great power, which slides comfortably into the surrounding vocabulary and is what most people assume they are singing. The difference is one letter and it changes the name entirely. Where the reading with the dental is kept, it is generally explained as her attachment to Śiva, and read that way it is a pointed thing to place here: eleven names have described the untying of every knot that binds, and the twelfth affirms one bond that is not untied. The tradition's technical term for what the yogin must abandon is turned into a name of the goddess. It is worth saying plainly that this is a real textual disagreement rather than a matter of interpretation, and that a reader will hear both readings from teachers of equal standing.

Kuṇḍalinī is not a compound. It is kuṇḍala, a ring or a coil or the coiled ornament worn in the ear, with the possessive suffix -in and its feminine -ī: the one who has a coil, the coiled one. The picture behind the word is a snake asleep at the base of the spine, wound three and a half times around, with its mouth stopping the opening of the central channel — so that the same body that blocks the road is the traveller who will take it. Eleven names of ascent have gone by before the word is spoken, and when it comes it is the shortest name in the passage, four syllables in the middle of a line, given no emphasis at all.

Bisa is the stalk of the lotus, and if you break one the pith draws out into filaments that hang between the halves before they snap. Sanskrit poets used those threads as the standard measure of something at the edge of perception but still continuous, still one thing. Tanīyasī is the comparative of tanu, slender, so the name says finer than that. The claim is about subtlety rather than smallness — the same order of statement as the Śvetāśvatara's description of the individual soul as a fraction of a fraction of the tip of a hair, which nonetheless has infinity in it. And the placement matters: the verse that began with lightning ends by saying she is too fine to see. Neither image cancels the other, and the passage as a whole seems designed to prevent the reciter from settling into either.

The pattern of the two preceding verses breaks here. There, four names filled four quarters exactly; here five names occupy the same thirty-two syllables, and the second line divides four, four, eight — mahāsaktiḥ, kuṇḍalinī, bisatantutanīyasī. Reciters carrying momentum from the long names of the ascent tend to crowd the two short ones together into something that sounds like a single word. Give each of the four-syllable names its own space. The name of the whole subject of the passage is one of them.

Two names in this verse end in visarga: taḍillatāsamaruciḥ and mahāsaktiḥ. Visarga is not silence and not an English h. It is a voiceless breath that echoes the vowel before it, so ruciḥ closes with a faint unvoiced i after the h, something like ru-chih(i), and saktiḥ likewise. Dropping it is the single most common fault in recitation of this hymn, and it matters more here than usual because in both cases the visarga is the grammatical marker that identifies these as nominatives of i-stems rather than the ā-stems that surround them.

Ṣaṭcakropari opens with a cluster that gives trouble in every language: the retroflex ṭ immediately followed by the palatal c. The tongue-tip is curled back for the first and must land at the hard palate for the second with no vowel between. The underlying numeral is ṣaṣ, six, whose final consonant becomes ṭ before the c of cakra; knowing that keeps you from inserting the vowel your mouth wants. Then kro carries a long o, which in Sanskrit is always long, and pari ends the compound before saṃ begins — the anusvāra there is a nasal made in the position of the following sth, so it is closer to n than to m.

Kuṇḍalinī has ṇḍ, two retroflexes in succession, and the temptation is to make the first one dental because English has no other kind of n before d. Curl the tongue back once and hold it for both. Taḍillatā gives you a genuine doubled l, produced by the sandhi of taḍit and latā, and a doubled consonant in Sanskrit is held twice as long: ta-ḍil-la-tā, with the l audibly sustained across the syllable break.

This verse is unusually instructive for the archanā forms, because its five names decline in four different ways. The ā-stem takes -āyai: ṣaṭcakroparisaṃsthitāyai namaḥ. The two i-stems take -yai on a stem whose final i becomes a glide: taṭillatāsamarucyai namaḥ, mahāsaktyai namaḥ. The two ī-stems take -yai likewise but with the long vowel shortening to the glide: kuṇḍalinyai namaḥ, bisatantutanīyasyai namaḥ. Working through all five once slowly does more for the feminine dative than memorising a paradigm.

As for what the verse asks: it offers two pictures of her and refuses to reconcile them, and it gives no instruction whatever. Nobody is told to raise anything, look for lightning, or expect a coil to stir. The names are descriptions in the third person of what is the case, and a reciter who treats them as a report of what should be happening in his own spine has mistaken the grammar. The one thing the verse does ask, by its metre, is that the name kuṇḍalinī not be hurried.